feat(mail): adapt M365 ETL loop for OO+Gmail bots.

Default compose mail-sync to onlyoffice,gmail every 300s with import on new
mail; gate full --rebuild behind MAIL_SYNC_INDEX. Fix mail-build stage order
and add bin/stack/start-mail-sync.

.gitignore

@@ -14,3 +14,9 @@ go.work.local
 models/
 # Purged from git history. Do not re-add.
 docs/crm-associations-proof.md
+
+# mount scaffold for the 8TB volume, never part of the repo
+mnt/
+
+# go build ./bin/mail/sync.go drops a binary named `sync` in cwd
+/sync

AGENTS.md

@@ -48,11 +48,12 @@ bin/postgres/ query.go (read-only YAML)
 bin/git/      import.go (go-git history; Python shim execs it)
 bin/web/      search.go (SearXNG; Python shim execs it)
 bin/reasoner/ bakeoff.go (D18 CPU OpenAI tool-call bake-off)
-internal/     shared Go (brain/rank is cgo-free; chats parsers; gitlog; websearch; reasoner; duckstats)
+internal/     shared Go (brain/rank is cgo-free; facts D16; cli flaggy D23; chats; gitlog; websearch; reasoner; duckstats)
 bin/qa/       stats.go (DuckDB quantiles / JSONL count; gcc CGO, not Zig)
 bin/watch/    corpus watcher (used by bin/brain/watch.go)
 bin/tools/    vendored python libs behind bin/* (kblib, yamlout, websearch)
 bin/cgo/      zig zcc zc++ (CGO via zig cc, not gcc)
+bin/stack/    start start-assistant stop status (compose + PicoClaw agent)
 bin/docker-entrypoint  container entrypoint (api: serve|search|watch; index: python)
 compose.yaml  docker composition (root level, not docker/)
 Dockerfile    api (Zig CGO, no Python) + index (Python write)
@@ -65,12 +66,20 @@ var/          kb.lbug, var/mail/*, caches (gitignored)
 ```bash
 bin/mail/sync.go --source onlyoffice,gmail --workers 8 --out var/mail   # raw message.json + attachments
 bin/mail/sync.go --source gmail --query 'from:example.com' --out var/mail  # Gmail search (default in:inbox)
+bin/mail/sync.go --source m365 --env ~/.config/brain/mail.env --out var/mail  # Microsoft Graph (delta)
 bin/mail/import.go --from-raw var/mail                                  # message.json → message.md (convert only)
 bin/brain/index.go --rebuild --with-facts --with-chats
+bin/stack/start-mail-sync                                               # compose ETL: sync→import every 300s
 ```
 
 - `sync` (Go) downloads messages + attachments; Gmail uses paginated list +
-  `body.attachmentId` (not partId) for attachments.
+  `body.attachmentId` (not partId) for attachments. Sources: `onlyoffice`,
+  `gmail`, `m365` (client-credentials + delta link; commit after success).
+- Compose `mail-sync` / `bin/stack/start-mail-sync`: ETL loop (default
+  `onlyoffice,gmail`, 300s). On `new>0` runs import; full `--rebuild` only if
+  `MAIL_SYNC_INDEX=1`. Secrets: `~/.config/brain/mail.env` + `~/.gmail-mcp`.
+  Case wrappers (e.g. family `gmail-sync-la-quinta.sh`) and ai-bot
+  `gmail-reauth.sh` reuse this sync/OAuth — do not fork corpus download.
 - `import` converts body + attachments to markdown. PDFs use poppler
   `pdftotext -layout` fast path (~15ms); textless/scanned PDFs use
   `pdftoppm` + tesseract `eng+deu` (`bin/mail/ocr.go`). Optional
@@ -84,11 +93,13 @@ bin/brain/index.go --rebuild --with-facts --with-chats
 ## Tools
 
 ```bash
-bin/facts/audit.go ["self"|"facts"|"info"|"stale"]  # 2-source + staleness gate
+bin/facts/audit.go ["self"|"db"|"contradict"]     # 2-source + D16 adjudication
 bin/facts/crm.go [--dry-run]                       # proof person↔company/company↔project (ooCRM × corpus SoT)
 bin/kb/search "query" [--repo X]                  # deprecated wrapper → bin/brain/search.go
 bin/brain/search.go "query" [--root facts|info]   # deduction search → YAML
+bin/brain/search.go "query" --as-of 2025-01-01    # D24 fact intervals
 bin/brain/search.go "query" --no-web              # local graph only
+source <(./bin/cli/complete.go bash)              # flaggy completions (D23)
 eval "$(bin/cgo/zig env)"                         # Zig cc + liblbug (not gcc)
 bin/brain/index.go --rebuild [--with-mail] [--with-facts] [--with-chats]
 bin/brain/add.go --text T --root facts --source "a.md x b.md"  # incremental write
@@ -97,6 +108,10 @@ bin/brain/get.go <id> [--body] [--json]          # Go read; Python bin/kb/get CI
 bin/brain/stats.go [--json]
 bin/brain/eval.go [--json]                       # recall@5; questions in internal/brain/rank
 bin/brain/serve.go                               # HTTP :8630; GET /openapi.json POST /mcp
+bin/stack/start                                  # brain HTTP/MCP (reuse healthy :8630)
+bin/stack/start-assistant                        # + reasoner + PicoClaw agent
+bin/stack/status                                 # YAML health
+bin/stack/stop                                   # compose stop; volumes kept
 bin/markdown/import.go [dir]                      # H2 leafs → YAML; Python bin/md/import fallback
 bin/git/import.go [REPO] [--json] [--limit N]     # go-git history → commit leafs
 bin/web/search.go "query" [--json]                # SearXNG; throttled ≠ absence

Dockerfile

@@ -6,6 +6,15 @@
 # API: Go + ladybug via Zig CGO (no CPython).
 # Index: Python write path (profile `index` until brain/add is v2).
 
+# --- mail-sync: standalone M365/OnlyOffice/Gmail puller (pure Go, no CGO) ---
+FROM golang:1.26-bookworm AS mail-build
+WORKDIR /src
+COPY go.mod go.sum ./
+RUN go mod download
+COPY bin/mail ./bin/mail
+COPY internal ./internal
+RUN CGO_ENABLED=0 go build -o /mail-sync ./bin/mail/sync.go
+
 # --- Python sidecar (Ladybug write / rebuild) ---
 FROM python:3.12-slim AS index
 
@@ -28,6 +37,7 @@ RUN python -m pip install --no-cache-dir -r /tmp/requirements.lock.txt \
 COPY . .
 RUN chmod +x /app/bin/docker-entrypoint \
     && chown -R 2dph:2dph /app
+COPY --from=mail-build /mail-sync /app/bin/mail-sync
 USER 2dph
 
 ENV PATH="/app/bin:${PATH}" \

PLAN.md

@@ -5,8 +5,12 @@ RAG over the operational Brain/ops/eSlider stack. Built like Sherlock
 Holmes: nothing is asserted unless it has proof.
 
 Status: **v1 in** (epic [#16](https://git.produktor.io/eSlider/2dph/issues/16) closed).
-v2 board: milestone [v2](https://git.produktor.io/eSlider/2dph/milestone/13) — OCR [#6](https://git.produktor.io/eSlider/2dph/issues/6),
+v2 board: milestone [v2](https://git.produktor.io/eSlider/2dph/milestone/13) —
-[#29](https://git.produktor.io/eSlider/2dph/issues/29) OQ1, [#30](https://git.produktor.io/eSlider/2dph/issues/30) OQ3.
+OCR [#6](https://git.produktor.io/eSlider/2dph/issues/6) in,
+[#29](https://git.produktor.io/eSlider/2dph/issues/29) OQ1 in,
+[#30](https://git.produktor.io/eSlider/2dph/issues/30) OQ3 in,
+[#34](https://git.produktor.io/eSlider/2dph/issues/34) D23 in,
+[#36](https://git.produktor.io/eSlider/2dph/issues/36) OQ5/D24 in.
 Gap: [docs/roadmap.md](docs/roadmap.md).
 
 ## What
@@ -42,13 +46,15 @@ detective method: **a fact needs ≥2 independent sources or it is
 | D13 | portfolio | start graph `(Person:eslider)-[:HAS]->(Portfolio)`, associate other natural/juristic persons later. |
 | D14 | tooling style | `bin/{subject}/{method}.go` shebang (e.g. `bin/brain/search.go`). Shared code in `internal/`. One root `go.mod` + `go.work`. No `bin/*/main.go`, no nested modules. |
 | D15 | repo | Gitea [`eSlider/2dph`](https://git.produktor.io/eSlider/2dph) is origin + [issues](https://git.produktor.io/eSlider/2dph/issues). GitHub `eSlider/2dph` is the public clone (PRs + Actions CI). No direct `main` pushes. TDD → PR → CI green → merge. |
-| D16 | contradictions | ≥2 yes vs ≥2 no → unrelated sources conflict → hypothesis → `(not confirmed)`. Resolution (authority, staleness adjudication) = **v2**, tracked as open question. |
+| D16 | contradictions | ≥2 yes vs ≥2 no → hypothesis → `(not confirmed)` until a rule fires. Order: **temporal_freshness** (fresh ≥2 vs stale minority), then **authority_pairing** (runtime/config A×B beats narrative C). Store as `a x b vs c x d` on hypothesis leafs. `bin/facts/audit contradict`. [#29](https://git.produktor.io/eSlider/2dph/issues/29). |
 | D17 | assertion gate | Fact-check every *claim* (facts → info → live → web), not every edit. `bin/brain/search.go` adds a `web` block when there is no facts hit (`throttled`/`skipped`/`refused` ≠ absence). `--root` and `--no-web` stay local. Missing graph ≠ “does not exist”. |
 | D18 | reasoner | Pluggable OpenAI-compatible URL (`REASONER_BASE_URL`). RAM: `Qwen/Qwen3.5-9B`. Quality: `prism-ml/Bonsai-27B-gguf` or `Qwen/Qwen3.6-27B`. No official Qwen3.6-9B. CPU bake-off: `bin/reasoner/bakeoff.go` + compose profile `reasoner` (`OLLAMA_NUM_GPU=0`, `:11435`). PicoClaw is compose profile `picoclaw`; tools are `search`/`get`/`audit`. Weights are not copied into the 2dph image. Agent lever/loop: [#15](https://git.produktor.io/eSlider/2dph/issues/15). |
 | D19 | git history | [go-git](https://github.com/go-git/go-git) via `bin/git/import.go`. No subprocess of the git binary. Conversion prints commit leafs; brain write is `bin/brain/index.go`. |
 | D20 | agent API | OpenAPI + MCP are generated from the same `internal/httpapi.Ops` table as `bin/brain/serve.go` handlers. `GET /openapi.json`, `POST /mcp` (JSON-RPC tools/list + tools/call). Tool names match OpenAPI paths (`search`/`get`/`stats`/`audit`/`ingest`). |
 | D21 | CGO | Ladybug/tokenizers CGO is compiled with **Zig** (`bin/cgo/zcc` → `zig cc -target …-linux-gnu`), not gcc. `bin/cgo/zig` pins Zig 0.14.1 + liblbug 0.19.1 + libtokenizers 1.27.0. Compose `target: api` has no CPython; write/rebuild is profile `index`. |
 | D22 | analytics | **duckdb-go** in-process (`internal/duckstats`, `bin/qa/stats.go`) for quantiles/JSONL. Links with **gcc/g++**, not Zig. Ladybug stays the graph; web-search cache stays modernc sqlite. Slice small structured docs with **mikefarah/yq**, not kislyuk/jq. [#30](https://git.produktor.io/eSlider/2dph/issues/30). |
+| D23 | CLI | **flaggy** (`github.com/integrii/flaggy`, 0 deps). Flags at any position. Wrapper `internal/cli`. Bash complete: `source <(./bin/cli/complete.go bash)`. No cobra, no stdlib `flag` in Go tools. Search does not intercept the word `completion`. [#34](https://git.produktor.io/eSlider/2dph/issues/34). |
+| D24 | fact intervals | Leaf `valid_from` / `valid_to` (YYYY-MM-DD, inclusive; empty = open/legacy). Search `--as-of` / MCP `as_of` keeps facts active that day. Not D16 `temporal_freshness` (source stale vs HEAD). Empty interval = always visible. [#36](https://git.produktor.io/eSlider/2dph/issues/36). |
 
 ## Architecture
 
@@ -65,7 +71,8 @@ detective method: **a fact needs ≥2 independent sources or it is
     brain/add.go            incremental leaf write (no rebuild)
     brain/get.go stats.go eval.go  # Go read (cgo); Python bin/kb/* CI fallback
     brain/watch.go
-    brain/search.go         deduction: facts → info → web-search
+    brain/search.go         deduction: facts → info → web
+    cli/complete.go         flaggy bash/zsh/fish complete (D23)
     brain/serve.go          HTTP API in-process + OpenAPI/MCP (D20); Zig CGO (D21)
     cgo/zig zcc zc++        CGO toolchain (zig cc, not gcc)
     mail/import.go          JSON → markdown (no brain write)
@@ -78,7 +85,8 @@ detective method: **a fact needs ≥2 independent sources or it is
                             (libs in internal/chats; no chats index)
     mail/ocr.go             tesseract eng+deu (pdftoppm scans)
     md/import               (deprecated; bin/markdown/import.go)
-    brain/extract  brain/audit   brain/deduce    (thinking wrapper)
+    brain/extract  brain/audit       brain/deduce    (thinking wrapper)
+    stack/start start-assistant start-mail-sync stop status
     web/search               (deprecated shim → web/search.go)
     db/psql-yq               (vendored)
     ssh-tunnel               onlyoffice pg tunnel 5433
@@ -96,7 +104,8 @@ them from each hit (1=File, 2=Commit, 3=Person). Rebuild writes
 `Leaf-[:FROM_FILE]->File`; git import writes the rest.
 
 Common props on every node/edge: `root`, `confidence`, `evidence[]`, `how`,
-`where`, `when`, `source_rev`.
+`where`, `when`, `source_rev`. Leaf interval of truth (D24): `valid_from`,
+`valid_to`.
 
 ## Config
 
@@ -117,8 +126,8 @@ Common props on every node/edge: `root`, `confidence`, `evidence[]`, `how`,
 
 ## Open questions (v2)
 
-- OQ1: mutually-contradicting evidence — how to resolve (authority weighting,
+- OQ1: **in** — D16 adjudication: `temporal_freshness` then `authority_pairing`.
-  temporal freshness, audit adjudication). **v2**; [#29](https://git.produktor.io/eSlider/2dph/issues/29).
+  Unresolved 2v2 stays hypothesis. [#29](https://git.produktor.io/eSlider/2dph/issues/29).
 - OQ2: OCR — **in**. `pdftotext -layout` first; scans `pdftoppm` + tesseract
   `eng+deu` (`bin/mail/ocr.go`, `internal/ocr`). No gocv, no gosseract CGO
   (D21 Zig owns Ladybug CGO). Optional `OCR_ENGINE=paddle` / compose profile
@@ -127,11 +136,14 @@ Common props on every node/edge: `root`, `confidence`, `evidence[]`, `how`,
   quantiles / JSONL count. Not a second graph. [#30](https://git.produktor.io/eSlider/2dph/issues/30).
 - OQ4: YAML-first storage for leafs — deferred: JSON is ~10x faster to
   serialize and unambiguous; YAML only where humans edit files.
+- OQ5: **in** — fact `valid_from` / `valid_to` + `--as-of` / MCP `as_of` (D24).
+  Not D16 `temporal_freshness`. [#36](https://git.produktor.io/eSlider/2dph/issues/36).
 
 ## Mail pipeline (done)
 
-1. `bin/mail/sync.go` (Go, 8 workers) — paginated Gmail/OnlyOffice download.
+1. `bin/mail/sync.go` (Go, 8 workers) — paginated Gmail / OnlyOffice / M365
-   Gmail attachments key off `body.attachmentId`, not MIME `partId`.
+   Graph download. Gmail attachments key off `body.attachmentId`, not MIME
+   `partId`. M365 uses client-credentials + delta link (commit after success).
 2. `bin/mail/import.go --from-raw` — message.json → message.md; PDFs via
    `pdftotext -layout` (~15ms); textless/scanned PDFs `pdftoppm` + tesseract
    `eng+deu`. ICS sidecars
@@ -140,7 +152,11 @@ Common props on every node/edge: `root`, `confidence`, `evidence[]`, `how`,
    corrupts its WAL on bulk-insert into an already-indexed DB. Conversion and
    indexing stay separate for crash safety. `bin/mail/index_mail` is a
    deprecation shim.
-4. Result: 17,835 messages → 28,918 info leafs, FTS + HNSW healthy, searchable
+4. Compose `mail-sync` / `bin/stack/start-mail-sync` — ETL loop (default
+   `onlyoffice,gmail`, 300s): sync → import on `new>0`; full rebuild only if
+   `MAIL_SYNC_INDEX=1`. Bot digests (ai-bot) and case wrappers reuse sync/OAuth;
+   they do not replace the corpus path.
+5. Result: 17,835 messages → 28,918 info leafs, FTS + HNSW healthy, searchable
    via `bin/brain/search.go`.
 
 ## CI/CD pipeline (D15)
@@ -184,4 +200,4 @@ Narrative: [docs/roadmap.md](docs/roadmap.md).
 | 4 | [#15](https://git.produktor.io/eSlider/2dph/issues/15) | **in** — lever/loop documented (`search` → `get` → `audit`). |
 | 5 | [#19](https://git.produktor.io/eSlider/2dph/issues/19) | **in** — CI recall SoT is `bin/brain/eval.go` via Zig. Python `bin/kb/eval` stays as an explicit fallback. |
 
-Does **not** block epic close: OQ1 [#29](https://git.produktor.io/eSlider/2dph/issues/29), OQ3 [#30](https://git.produktor.io/eSlider/2dph/issues/30), OQ4. OCR [#6](https://git.produktor.io/eSlider/2dph/issues/6) is **in**.
+Does **not** block epic close: OQ4. OCR [#6](https://git.produktor.io/eSlider/2dph/issues/6), OQ1 [#29](https://git.produktor.io/eSlider/2dph/issues/29), OQ3 [#30](https://git.produktor.io/eSlider/2dph/issues/30) are **in**.

README.md

@@ -119,6 +119,8 @@ Mail is a first-class corpus (retrievable through the same search):
 
 ```bash
 bin/mail/sync.go --source onlyoffice,gmail --workers 8 --out var/mail  # raw sync (Go)
+bin/mail/sync.go --source m365 --env ~/.config/brain/mail.env          # Microsoft 365 Graph
+bin/stack/start-mail-sync                                              # compose ETL (300s; no auto-rebuild)
 bin/mail/import.go --from-raw var/mail                                  # JSON → markdown
 bin/brain/add.go --text T --root facts --source "a.md x b.md"
 bin/brain/index.go --rebuild --with-facts --with-chats                  # facts extract + chats md
@@ -160,6 +162,10 @@ go test ./... && uv run python -m unittest discover -s bin/tools -t .
 Docker (optional, cached model + var volumes):
 
 ```bash
+bin/stack/start                                # brain HTTP/MCP :8630
+bin/stack/start-assistant                      # + qwen3.5:9b + PicoClaw agent
+bin/stack/status
+bin/stack/stop
 docker compose up -d brain                     # API (Zig CGO serve :8630)
 docker compose --profile index run --rm index  # Python Ladybug rebuild
 docker compose --profile picoclaw up brain-mcp # MCP on 127.0.0.1:8630

bin/cli/complete.go

@@ -0,0 +1,91 @@
+//usr/bin/env go run "$0" "$@"; exit
+//
+// bin/cli/complete.go - dump flaggy shell completions for all Go shebang tools (D23).
+//
+//	source <(./bin/cli/complete.go bash)
+//	./bin/cli/complete.go zsh|fish|powershell|nushell
+//
+// Search does not steal the word "completion"; this binary dumps scripts.
+// NOTE: never run `gofmt -w` on this file — it breaks the shebang.
+package main
+
+import (
+	"fmt"
+	"os"
+	"strings"
+
+	mailsync "github.com/eSlider/2dph/bin/mail/sync"
+	"github.com/eSlider/2dph/internal/brain/rank"
+	"github.com/eSlider/2dph/internal/chats"
+	"github.com/eSlider/2dph/internal/cli"
+	"github.com/eSlider/2dph/internal/gitlog"
+	"github.com/eSlider/2dph/internal/mdleaves"
+	"github.com/eSlider/2dph/internal/ocr"
+	"github.com/eSlider/2dph/internal/reasoner"
+	"github.com/eSlider/2dph/internal/websearch"
+	"github.com/integrii/flaggy"
+)
+
+func tools() []cli.Tool {
+	return []cli.Tool{
+		{Path: "bin/brain/search.go", Name: "brain-search", New: rank.Parser},
+		{Path: "bin/brain/get.go", Name: "brain-get", New: func() *flaggy.Parser {
+			o := rank.GetOptions{}
+			return rank.GetParser(&o)
+		}},
+		{Path: "bin/brain/stats.go", Name: "brain-stats", New: rank.StatsParser},
+		{Path: "bin/brain/eval.go", Name: "brain-eval", New: rank.EvalParser},
+		{Path: "bin/web/search.go", Name: "web-search", New: websearch.Parser},
+		{Path: "bin/git/import.go", Name: "git-import", New: gitlog.Parser},
+		{Path: "bin/markdown/import.go", Name: "markdown-import", New: mdleaves.Parser},
+		{Path: "bin/qa/stats.go", Name: "qa-stats", New: cli.QAParser},
+		{Path: "bin/reasoner/bakeoff.go", Name: "reasoner-bakeoff", New: reasoner.Parser},
+		{Path: "bin/mail/ocr.go", Name: "mail-ocr", New: ocr.Parser},
+		{Path: "bin/mail/sync.go", Name: "mail-sync", New: mailsync.Parser},
+		{Path: "bin/chats/sync.go", Name: "chats-sync", New: chats.SyncParser},
+		{Path: "bin/chats/import.go", Name: "chats-import", New: chats.ImportParser},
+		{Path: "bin/chats/facts.go", Name: "chats-facts", New: chats.FactsParser},
+		{Path: "bin/chats/apply.go", Name: "chats-apply", New: chats.ApplyParser},
+	}
+}
+
+func main() {
+	os.Exit(run(os.Args[1:]))
+}
+
+func run(args []string) int {
+	shell := "bash"
+	if len(args) > 0 {
+		switch args[0] {
+		case "bash", "zsh", "fish", "powershell", "nushell":
+			shell = args[0]
+		case "-h", "--help", "help":
+			fmt.Fprintln(os.Stderr, "usage: bin/cli/complete.go [bash|zsh|fish|powershell|nushell]")
+			return 0
+		default:
+			fmt.Fprintf(os.Stderr, "cli/complete: unknown shell %q\n", args[0])
+			return 2
+		}
+	}
+	if shell == "bash" {
+		fmt.Print(cli.BashScript(tools()))
+		return 0
+	}
+	var b strings.Builder
+	for _, t := range tools() {
+		p := t.New()
+		p.Name = t.Name
+		switch shell {
+		case "zsh":
+			b.WriteString(flaggy.GenerateZshCompletion(p))
+		case "fish":
+			b.WriteString(flaggy.GenerateFishCompletion(p))
+		case "powershell":
+			b.WriteString(flaggy.GeneratePowerShellCompletion(p))
+		case "nushell":
+			b.WriteString(flaggy.GenerateNushellCompletion(p))
+		}
+	}
+	fmt.Print(b.String())
+	return 0
+}

bin/docker-entrypoint

@@ -5,6 +5,7 @@
 #   serve | search | watch
 # Index image (Python write path, compose profile `index`):
 #   index | extract | audit | search (deprecated python wrapper)
+#   mail-sync [N]   ETL loop: sync -> import; optional index (default 300s)
 #
 # Usage comment starts at line 2 (self-describing convention).
 set -euo pipefail
@@ -34,5 +35,27 @@ case "$CMD" in
 	serve) exec /app/bin/serve "$@" ;;
 	extract) exec "$KB_PY" /app/bin/facts/extract "$@" ;;
 	audit) exec "$KB_PY" /app/bin/facts/audit "$@" ;;
+	mail-sync)
+		# ETL: pull mail, convert to md when new>0. Full --rebuild is opt-in
+		# (MAIL_SYNC_INDEX=1) — ~29k leaf rebuild is minutes, not a 10s loop.
+		interval="${1:-300}"
+		[ "$interval" -gt 0 ] 2>/dev/null || interval=300
+		: "${MAIL_SYNC_ENV:=/secret/mail.env}"
+		: "${MAIL_SYNC_SRC:=onlyoffice,gmail}"
+		: "${MAIL_SYNC_OUT:=/app/var/mail}"
+		: "${MAIL_SYNC_INDEX:=0}"
+		while true; do
+			out="$("/app/bin/mail-sync" --source "$MAIL_SYNC_SRC" --env "$MAIL_SYNC_ENV" --out "$MAIL_SYNC_OUT" 2>&1)" || true
+			echo "$out"
+			new="$(printf '%s\n' "$out" | sed -n 's/.*new=\([0-9]*\).*/\1/p' | tail -1)"
+			if [ -n "$new" ] && [ "$new" -gt 0 ] 2>/dev/null; then
+				"$KB_PY" /app/bin/mail/import --from-raw "$MAIL_SYNC_OUT" 2>&1 | tail -1
+				if [ "$MAIL_SYNC_INDEX" = "1" ]; then
+					"$KB_PY" /app/bin/kb/index --rebuild --with-mail 2>&1 | tail -1
+				fi
+			fi
+			sleep "$interval"
+		done
+		;;
 	*) echo "unknown command: $CMD" >&2; exit 2 ;;
 esac

bin/facts/audit

@@ -1,14 +1,15 @@
 #!/usr/bin/env python3
 """facts/audit - evidence & lexicon checks for the 2dph brain.
 
-  bin/facts/audit self      # lexicon: every fact in db has >=2 sources
+  bin/facts/audit self         # lexicon: docs + two-source rule
-  bin/facts/audit db        # evidence gate: run against var/kb.lbug
+  bin/facts/audit db           # evidence gate against var/kb.lbug
+  bin/facts/audit contradict   # D16 adjudication (JSON claim(s) on stdin)
 
-`self` mode checks the repo itself (no network, no runtime deps). It greps
+`self` mode checks the repo itself (no network, no runtime deps).
-for known-good two-source pairings and confirms the docs are consistent.
+`db` mode loads every Leaf with root=facts. Confirmed facts need ` x `;
-`db` mode loads every Leaf with root=facts and asserts each has source_rev
+hypothesis contradictions need `a x b vs c x d` (both sides ≥2).
-and a non-empty `loc` (the "where did you see it" evidence pointer) and that
+`contradict` applies temporal_freshness then authority_pairing; ≥2 vs ≥2
-'confirmed' facts carry a two-source `source` field.
+with no rule stays hypothesis / `(not confirmed)`.
 
 Exit 0 = all checks pass, 1 = audit failures, 2 = could not evaluate.
 """
@@ -22,6 +23,8 @@ from pathlib import Path
 ROOT = Path(__file__).resolve().parents[2]
 sys.path.insert(0, str(ROOT / "bin" / "tools"))
 
+from contradict import adjudicate, check_fact_row  # noqa: E402
+
 
 def audit_db() -> list[str]:
     from kblib import connect
@@ -33,14 +36,8 @@ def audit_db() -> list[str]:
     r = conn.execute("MATCH (l:Leaf {root:'facts'}) RETURN l.id, l.source, l.loc, l.how, l.confidence")
     problems: list[str] = []
     for lid, source, loc, how, conf in r.get_all():
-        if conf != "confirmed":
+        problems.extend(check_fact_row(str(lid), str(source or ""), str(loc or ""),
-            problems.append(f"{lid}: facts require confidence='confirmed', got '{conf}'")
+                                       str(how or ""), str(conf or "")))
-        if not source or " x " not in source:
-            problems.append(f"{lid}: needs 2-source evidence in source, got '{source}'")
-        if not loc:
-            problems.append(f"{lid}: missing loc (evidence pointer)")
-        if not how:
-            problems.append(f"{lid}: missing how")
     conn.close()
     db.close()
     return problems
@@ -54,20 +51,50 @@ def audit_self() -> list[str]:
         problems.append("PLAN.md missing recall@5 gate")
     if re.search(r"(?i)facts must have.*2 sources|2.source", plan) is None:
         problems.append("PLAN.md missing the two-source evidence rule for facts")
+    if "temporal_freshness" not in plan or "authority_pairing" not in plan:
+        problems.append("PLAN.md missing D16 adjudication rules")
     if re.search(r"(?i)HNSW|BM25|deduction", (ROOT / "README.md").read_text()) is None:
         problems.append("README.md missing search/retrieval description")
     return problems
 
 
+def audit_contradict(raw: str) -> tuple[list[str], list[dict]]:
+    raw = raw.strip()
+    if not raw:
+        return ["contradict: empty stdin (JSON claim or {claims:[...]})"], []
+    try:
+        payload = json.loads(raw)
+    except json.JSONDecodeError as e:
+        return [f"contradict: invalid JSON: {e}"], []
+    if isinstance(payload, dict) and "claims" in payload:
+        claims = list(payload.get("claims") or [])
+    elif isinstance(payload, dict):
+        claims = [payload]
+    elif isinstance(payload, list):
+        claims = payload
+    else:
+        return ["contradict: expected object or list"], []
+    details = [adjudicate(c) for c in claims]
+    return [], details
+
+
 def main(argv: list[str]) -> int:
     import argparse
     p = argparse.ArgumentParser(description="evidence & lexicon audit")
-    p.add_argument("mode", choices=("self", "db"))
+    p.add_argument("mode", choices=("self", "db", "contradict"))
     p.add_argument("--json", action="store_true")
     a = p.parse_args(argv)
 
-    problems = audit_self() if a.mode == "self" else audit_db()
+    details: list[dict] = []
-    out = {"mode": a.mode, "ok": not problems, "problems": problems}
+    if a.mode == "self":
+        problems = audit_self()
+    elif a.mode == "db":
+        problems = audit_db()
+    else:
+        problems, details = audit_contradict(sys.stdin.read())
+    out: dict = {"mode": a.mode, "ok": not problems, "problems": problems}
+    if details:
+        out["contradictions"] = details
     if a.json:
         print(json.dumps(out, indent=2))
     else:

bin/facts/audit.go

@@ -5,6 +5,7 @@
 //
 //	./bin/facts/audit.go self
 //	./bin/facts/audit.go db
+//	./bin/facts/audit.go contradict --json < claim.json
 //
 // Python bin/facts/audit is the implementation (CI runs it directly).
 // NOTE: never run `gofmt -w` on this file — it breaks the shebang.

bin/git/import.go

@@ -16,9 +16,8 @@ import (
 	"fmt"
 	"os"
 	"path/filepath"
-	"strconv"
-	"time"
 
+	cliparse "github.com/eSlider/2dph/internal/cli"
 	"github.com/eSlider/2dph/internal/cmdbin"
 	"github.com/eSlider/2dph/internal/gitlog"
 )
@@ -28,49 +27,16 @@ func main() {
 }
 
 func run(args []string) int {
-	var repo, root, since string
+	c, err := gitlog.ParseArgs(args)
-	limit := 0
+	if err != nil {
-	jsonOut := false
+		return cliparse.Fail(err)
-	i := 0
-	for i < len(args) {
-		a := args[i]
-		switch {
-		case a == "--json":
-			jsonOut = true
-		case a == "--limit" && i+1 < len(args):
-			i++
-			n, err := strconv.Atoi(args[i])
-			if err != nil || n < 0 {
-				fmt.Fprintf(os.Stderr, "git/import: --limit must be a non-negative integer\n")
-				return 2
-			}
-			limit = n
-		case a == "--since" && i+1 < len(args):
-			i++
-			since = args[i]
-		case a == "--root" && i+1 < len(args):
-			i++
-			root = args[i]
-		case a == "-h" || a == "--help":
-			fmt.Fprintln(os.Stderr, `usage: bin/git/import.go [REPO] [--json] [--limit N] [--since DATE] [--root DIR]`)
-			return 0
-		case len(a) > 0 && a[0] != '-':
-			repo = a
-		default:
-			fmt.Fprintf(os.Stderr, "git/import: unknown flag %s\n", a)
-			return 2
-		}
-		i++
 	}
+	repo, root, since, limit, jsonOut := c.Repo, c.Root, c.Since, c.Limit, c.JSONOut
 
-	var sinceT time.Time
+	sinceT, err := gitlog.ParseSince(since)
-	if since != "" {
+	if err != nil {
-		var err error
+		fmt.Fprintf(os.Stderr, "git/import: %v\n", err)
-		sinceT, err = parseSince(since)
+		return 2
-		if err != nil {
-			fmt.Fprintf(os.Stderr, "git/import: %v\n", err)
-			return 2
-		}
 	}
 
 	repos := []string{}
@@ -132,12 +98,3 @@ func run(args []string) int {
 	}
 	return 0
 }
-
-func parseSince(s string) (time.Time, error) {
-	for _, layout := range []string{time.RFC3339, "2006-01-02"} {
-		if t, err := time.Parse(layout, s); err == nil {
-			return t, nil
-		}
-	}
-	return time.Time{}, fmt.Errorf("cannot parse --since %q", s)
-}

bin/kb/add

@@ -65,6 +65,10 @@ def main(argv: list[str]) -> int:
     p.add_argument("--how", default="brain/add")
     p.add_argument("--loc", default="")
     p.add_argument("--type", default="reference", dest="type_")
+    p.add_argument("--valid-from", default="", dest="valid_from",
+                   help="fact interval start YYYY-MM-DD (D24)")
+    p.add_argument("--valid-to", default="", dest="valid_to",
+                   help="fact interval end YYYY-MM-DD inclusive; empty=open (D24)")
     args = p.parse_args(argv)
 
     if args.json:
@@ -86,6 +90,8 @@ def main(argv: list[str]) -> int:
             "how": args.how,
             "loc": args.loc or args.source,
             "type": args.type_,
+            "valid_from": args.valid_from,
+            "valid_to": args.valid_to,
         }]
 
     for lf in leafs:

bin/mail/ocr.go

@@ -17,6 +17,7 @@ import (
 	"os"
 	"strings"
 
+	cliparse "github.com/eSlider/2dph/internal/cli"
 	"github.com/eSlider/2dph/internal/ocr"
 )
 
@@ -25,15 +26,12 @@ func main() {
 }
 
 func run(args []string) int {
-	if len(args) != 1 || strings.HasPrefix(args[0], "-") {
+	c, err := ocr.ParseArgs(args)
-		fmt.Fprintln(os.Stderr, `usage: bin/mail/ocr.go <image|pdf>`)
+	if err != nil {
-		return 2
+		return cliparse.Fail(err)
 	}
-	path := args[0]
+	path := c.Path
-	var (
+	var text string
-		text string
-		err  error
-	)
 	if strings.HasSuffix(strings.ToLower(path), ".pdf") {
 		text, err = ocr.PDFFile(path)
 	} else {

bin/mail/sync.go

@@ -1,8 +1,9 @@
 //usr/bin/env go run "$0" "$@"; exit
-// bin/mail/sync.go - async download of OnlyOffice and Gmail mail to var/mail/.
+// bin/mail/sync.go - async download of OnlyOffice, Gmail and M365 mail to var/mail/.
 //
-//	./bin/mail/sync.go --source onlyoffice,gmail --limit 50 --workers 8
+//	./bin/mail/sync.go --source onlyoffice,gmail,m365 --limit 50 --workers 8
 //	./bin/mail/sync.go --source gmail --force
+//	./bin/mail/sync.go --source m365 --env .secrets/m365.env
 //	./bin/mail/sync.go --dry-run
 //
 // Writes raw message.json + attachments under var/mail/<folder>/<id>/; run

bin/mail/sync/cli.go

@@ -5,74 +5,103 @@ package sync
 
 import (
 	"context"
-	"flag"
+	"errors"
 	"fmt"
 	"os"
 	"path/filepath"
 	"strings"
 	"time"
+
+	cliparse "github.com/eSlider/2dph/internal/cli"
+	"github.com/integrii/flaggy"
 )
 
 // CLIConfig is a superset of SyncConfig plus flag parsing results.
 type CLIConfig struct {
-	Sync SyncConfig
+	Sync    SyncConfig
-	Env  string // .env path; default <cwd>/.env
+	Env     string // .env path; default <cwd>/.env
 	Sources string
-	Help bool
+	Help    bool
 }
 
-// ParseCLI reads os.Args into a CLIConfig. Exit codes: 0 ok, 2 usage.
+type flagVals struct {
+	env, out, srcs, query string
+	workers, limit, offset int
+	force, dryRun         bool
+}
+
+func Parser() *flaggy.Parser {
+	v := flagVals{workers: 4, query: "in:inbox", srcs: "onlyoffice"}
+	return bind(&v)
+}
+
+func bind(v *flagVals) *flaggy.Parser {
+	if v.workers == 0 {
+		v.workers = 4
+	}
+	if v.query == "" {
+		v.query = "in:inbox"
+	}
+	if v.srcs == "" {
+		v.srcs = "onlyoffice"
+	}
+	p := cliparse.New("mail-sync")
+	p.Description = "download mail to var/mail"
+	p.String(&v.env, "", "env", ".env file")
+	p.String(&v.out, "", "out", "var/mail root")
+	p.Int(&v.workers, "", "workers", "concurrent downloads")
+	p.Int(&v.limit, "", "limit", "max messages per source (0 = all)")
+	p.Int(&v.offset, "", "offset", "skip first N messages per source")
+	p.Bool(&v.force, "", "force", "overwrite existing message.json")
+	p.Bool(&v.dryRun, "", "dry-run", "list counts without writing")
+	p.String(&v.query, "", "query", "Gmail search query")
+	p.String(&v.srcs, "", "source", "comma list: onlyoffice,gmail,m365")
+	return p
+}
+
+// ParseCLI reads args into a CLIConfig. Exit codes: 0 ok, 2 usage.
 func ParseCLI(args []string) (CLIConfig, int, error) {
-	fs := flag.NewFlagSet("mail/sync", flag.ContinueOnError)
+	v := flagVals{workers: 4, query: "in:inbox", srcs: "onlyoffice"}
-	var (
+	p := bind(&v)
-		env    = fs.String("env", "", ".env file (default: <cwd>/.env)")
+	if err := cliparse.Parse(p, args); err != nil {
-		out    = fs.String("out", "", "var/mail root (default: <cwd>/var/mail)")
+		if errors.Is(err, cliparse.ErrHelp) {
-		workers = fs.Int("workers", 4, "concurrent downloads")
+			return CLIConfig{Help: true}, 0, nil
-		limit  = fs.Int("limit", 0, "max messages per source (0 = all)")
+		}
-		offset = fs.Int("offset", 0, "skip first N messages per source")
-		force  = fs.Bool("force", false, "overwrite existing message.json + attachments")
-		dryRun = fs.Bool("dry-run", false, "list message counts without writing")
-		query  = fs.String("query", "in:inbox", "Gmail search query (gmail source only)")
-		srcs   = fs.String("source", "onlyoffice", "comma list: onlyoffice,gmail (default onlyoffice)")
-		help   = fs.Bool("help", false, "usage")
-	)
-	fs.SetOutput(os.Stderr)
-	if err := fs.Parse(args); err != nil {
 		return CLIConfig{}, 2, err
 	}
-	if *help || fs.NArg() > 0 {
+	if len(p.TrailingArguments) > 0 {
 		return CLIConfig{Help: true}, 0, nil
 	}
 	wd, err := os.Getwd()
 	if err != nil {
 		return CLIConfig{}, 2, err
 	}
-	if *env == "" {
+	if v.env == "" {
-		*env = filepath.Join(wd, ".env")
+		v.env = filepath.Join(wd, ".env")
 	}
-	if *out == "" {
+	if v.out == "" {
-		*out = filepath.Join(wd, "var", "mail")
+		v.out = filepath.Join(wd, "var", "mail")
 	}
-	envVars := readEnv(*env)
+	envVars := readEnv(v.env)
 	cfg := SyncConfig{
-		Out:     *out,
+		Out:     v.out,
-		Workers: *workers,
+		Workers: v.workers,
-		Limit:   *limit,
+		Limit:   v.limit,
-		Offset:  *offset,
+		Offset:  v.offset,
-		Force:   *force,
+		Force:   v.force,
-		DryRun:  *dryRun,
+		DryRun:  v.dryRun,
-		Query:   *query,
+		Query:   v.query,
 		Policy:  RetryPolicy{},
 	}
-	cli := CLIConfig{Sync: cfg, Env: *env, Sources: *srcs}
+	out := CLIConfig{Sync: cfg, Env: v.env, Sources: v.srcs}
-	for _, s := range strings.Split(*srcs, ",") {
+	for _, s := range strings.Split(v.srcs, ",") {
 		switch strings.TrimSpace(s) {
 		case "onlyoffice":
 			u := pick(envVars["ONLYOFFICE_URL"], envVars["OO_URL"])
 			user := pick(envVars["ONLYOFFICE_USER"], envVars["OO_USER"])
 			pass := pick(envVars["ONLYOFFICE_PASS"], envVars["OO_PASSWORD"])
 			if u == "" || user == "" || pass == "" {
-				return CLIConfig{}, 2, fmt.Errorf("onlyoffice source needs ONLYOFFICE_URL/USER/PASS in %s", *env)
+				return CLIConfig{}, 2, fmt.Errorf("onlyoffice source needs ONLYOFFICE_URL/USER/PASS in %s", v.env)
 			}
 			cfg.OO = &OOConfig{URL: u, User: user, Password: pass}
 		case "gmail":
@@ -81,34 +110,52 @@ func ParseCLI(args []string) (CLIConfig, int, error) {
 				CredentialsPath: filepath.Join(home, ".gmail-mcp", "credentials.json"),
 				KeysPath:        filepath.Join(home, ".gmail-mcp", "gcp-oauth.keys.json"),
 			}
+		case "m365":
+			tenant := pick(envVars["M365_TENANT"], envVars["MS_TENANT"])
+			cid := pick(envVars["M365_CLIENT_ID"], envVars["MS_CLIENT_ID"])
+			sec := pick(envVars["M365_CLIENT_SECRET"], envVars["MS_CLIENT_SECRET"])
+			users := pick(envVars["M365_USERS"], envVars["MS_USERS"])
+			if tenant == "" || cid == "" || sec == "" || users == "" {
+				return CLIConfig{}, 2, fmt.Errorf("m365 source needs M365_TENANT/CLIENT_ID/CLIENT_SECRET/USERS in %s", v.env)
+			}
+			var userList []string
+			for _, u := range strings.Split(users, ",") {
+				if u = strings.TrimSpace(u); u != "" {
+					userList = append(userList, u)
+				}
+			}
+			if len(userList) == 0 {
+				return CLIConfig{}, 2, fmt.Errorf("m365 source: M365_USERS empty")
+			}
+			cfg.M365 = &M365Credentials{Tenant: tenant, ClientID: cid, ClientSecret: sec, Users: userList}
 		default:
 			return CLIConfig{}, 2, fmt.Errorf("unknown source %q", s)
 		}
 	}
-	cli.Sync = cfg
+	out.Sync = cfg
-	return cli, 0, nil
+	return out, 0, nil
 }
 
 // Main is the CLI entry: returns process exit code.
 func Main(args []string) int {
-	cli, code, err := ParseCLI(args)
+	cfg, code, err := ParseCLI(args)
 	if err != nil {
 		fmt.Fprintln(os.Stderr, "mail/sync:", err)
 		return code
 	}
-	if cli.Help {
+	if cfg.Help {
-		fmt.Fprintln(os.Stderr, "usage: bin/mail/sync.go [--source onlyoffice,gmail] [--query GMAIL_Q] [--limit N] [--offset N] [--workers N] [--force] [--dry-run]")
+		fmt.Fprintln(os.Stderr, "usage: bin/mail/sync.go [--source onlyoffice,gmail,m365] [--query GMAIL_Q] [--limit N] [--offset N] [--workers N] [--force] [--dry-run]")
 		return 0
 	}
 	ctx, cancel := context.WithTimeout(context.Background(), 6*time.Hour)
 	defer cancel()
 	start := time.Now()
-	stats, err := Run(ctx, cli.Sync)
+	stats, err := Run(ctx, cfg.Sync)
 	if err != nil {
 		fmt.Fprintln(os.Stderr, "mail/sync:", err)
 		return 1
 	}
-	if cli.Sync.DryRun {
+	if cfg.Sync.DryRun {
 		fmt.Printf("mail/sync: dry-run checked=%d (no writes)\n", stats.Checked)
 		return 0
 	}
@@ -135,13 +182,13 @@ func readEnv(path string) map[string]string {
 		k, v, _ := strings.Cut(line, "=")
 		out[strings.TrimSpace(k)] = strings.Trim(strings.TrimSpace(v), "\"'")
 	}
-	// env overrides file
 	for _, kv := range os.Environ() {
 		k, v, ok := strings.Cut(kv, "=")
 		if !ok {
 			continue
 		}
-		if strings.HasPrefix(k, "ONLYOFFICE_") || strings.HasPrefix(k, "OO_") {
+		if strings.HasPrefix(k, "ONLYOFFICE_") || strings.HasPrefix(k, "OO_") ||
+			strings.HasPrefix(k, "M365_") || strings.HasPrefix(k, "MS_") {
 			out[k] = v
 		}
 	}

bin/mail/sync/m365.go

@@ -0,0 +1,382 @@
+package sync
+
+import (
+	"context"
+	"encoding/json"
+	"errors"
+	"fmt"
+	"io"
+	"net/http"
+	"net/url"
+	"os"
+	"path/filepath"
+	"strings"
+	"time"
+)
+
+// M365Credentials holds a Microsoft Graph app registration with the Mail.Read
+// application permission. Client credentials are read from env/.env, never
+// committed.
+type M365Credentials struct {
+	Tenant       string
+	ClientID     string
+	ClientSecret string
+	Users        []string // mailbox addresses to sync, e.g. info@example.com
+}
+
+// m365Token is the cached access token with expiry.
+type m365Token struct {
+	AccessToken string
+	Expiry      time.Time
+}
+
+// M365Client talks to the Microsoft Graph API using the client-credentials
+// flow (app registration with Mail.Read application permission). GET-only:
+// messages are never marked as read or deleted.
+type M365Client struct {
+	creds         M365Credentials
+	base          string // graph base URL; default https://graph.microsoft.com
+	tokenEndpoint string // login endpoint; default https://login.microsoftonline.com
+	client        *http.Client
+	mu            chan struct{}
+	token         *m365Token
+}
+
+func NewM365Client(creds M365Credentials) (*M365Client, error) {
+	if creds.Tenant == "" || creds.ClientID == "" || creds.ClientSecret == "" {
+		return nil, errors.New("m365 needs tenant, client id and client secret")
+	}
+	c := &M365Client{
+		creds:  creds,
+		base:   "https://graph.microsoft.com",
+		client: &http.Client{Timeout: 90 * time.Second},
+		mu:     make(chan struct{}, 1),
+	}
+	c.mu <- struct{}{}
+	return c, nil
+}
+
+// accessToken returns a fresh bearer token, refreshing via the Azure AD token
+// endpoint when the cached one is missing or about to expire (within 2 min).
+func (c *M365Client) accessToken(ctx context.Context) (string, error) {
+	select {
+	case <-c.mu:
+	case <-ctx.Done():
+		return "", ctx.Err()
+	}
+	defer func() { c.mu <- struct{}{} }()
+	if c.token != nil && c.token.AccessToken != "" && time.Now().Before(c.token.Expiry.Add(-2*time.Minute)) {
+		return c.token.AccessToken, nil
+	}
+	return c.refreshLocked(ctx)
+}
+
+func (c *M365Client) refreshLocked(ctx context.Context) (string, error) {
+	form := url.Values{}
+	form.Set("grant_type", "client_credentials")
+	form.Set("client_id", c.creds.ClientID)
+	form.Set("client_secret", c.creds.ClientSecret)
+	form.Set("scope", "https://graph.microsoft.com/.default")
+
+	endpoint := c.tokenEndpoint
+	if endpoint == "" {
+		endpoint = fmt.Sprintf("https://login.microsoftonline.com/%s/oauth2/v2.0/token", c.creds.Tenant)
+	}
+	req, err := http.NewRequestWithContext(ctx, http.MethodPost, endpoint, strings.NewReader(form.Encode()))
+	if err != nil {
+		return "", err
+	}
+	req.Header.Set("Content-Type", "application/x-www-form-urlencoded")
+	resp, err := c.client.Do(req)
+	if err != nil {
+		return "", fmt.Errorf("m365 token: %w", err)
+	}
+	defer resp.Body.Close()
+	body, _ := io.ReadAll(io.LimitReader(resp.Body, 1<<20))
+	if resp.StatusCode != http.StatusOK {
+		var e struct {
+			Error string `json:"error"`
+			Desc  string `json:"error_description"`
+		}
+		_ = json.Unmarshal(body, &e)
+		return "", fmt.Errorf("m365 token status %d: %s (%s)", resp.StatusCode, e.Error, truncate(e.Desc, 200))
+	}
+	var out struct {
+		AccessToken string `json:"access_token"`
+		ExpiresIn   int64  `json:"expires_in"`
+	}
+	if err := json.Unmarshal(body, &out); err != nil {
+		return "", fmt.Errorf("m365 token parse: %w", err)
+	}
+	c.token = &m365Token{
+		AccessToken: out.AccessToken,
+		Expiry:      time.Now().Add(time.Duration(out.ExpiresIn) * time.Second),
+	}
+	return out.AccessToken, nil
+}
+
+// deltaPage is one response page of the Graph delta query.
+type deltaPage struct {
+	Value []struct {
+		ID            string `json:"id"`
+		RemovedReason string `json:"@odata.removedReason"`
+	} `json:"value"`
+	NextLink  string `json:"@odata.nextLink"`
+	DeltaLink string `json:"@odata.deltaLink"`
+}
+
+// ListDeltaIDs walks the inbox delta query and returns live message ids since
+// the previous deltaLink (or the full inbox when deltaLink is empty). Returns
+// the new deltaLink for the next run. GET-only; nothing is mutated server-side.
+func (c *M365Client) ListDeltaIDs(ctx context.Context, mailbox, deltaLink string, limit int) ([]string, string, error) {
+	var (
+		ids   []string
+		url   string
+		link  = deltaLink
+	)
+	if link == "" {
+		url = fmt.Sprintf("/v1.0/users/%s/mailFolders/inbox/messages/delta", pathEscape(mailbox))
+	} else {
+		url = link
+	}
+	for url != "" {
+		var page deltaPage
+		if err := c.getJSON(ctx, url, &page); err != nil {
+			return ids, link, err
+		}
+		for _, m := range page.Value {
+			if m.RemovedReason != "" {
+				continue
+			}
+			if m.ID == "" {
+				continue
+			}
+			ids = append(ids, m.ID)
+			if limit > 0 && len(ids) >= limit {
+				if page.DeltaLink != "" {
+					link = page.DeltaLink
+				}
+				return ids, link, nil
+			}
+		}
+		if page.DeltaLink != "" {
+			link = page.DeltaLink
+			url = ""
+			break
+		}
+		url = page.NextLink
+	}
+	return ids, link, nil
+}
+
+// GetMessage fetches a single message by id and normalizes it to the Message
+// contract. GET-only.
+func (c *M365Client) GetMessage(ctx context.Context, mailbox, id string) (*Message, error) {
+	path := fmt.Sprintf("/v1.0/users/%s/messages/%s?$expand=attachments($select=id,name,contentType,size,isInline)",
+		pathEscape(mailbox), pathEscape(id))
+	var raw struct {
+		ID                string             `json:"id"`
+		Subject           string             `json:"subject"`
+		From              m365Recipient      `json:"from"`
+		ToRecipients      []m365Recipient    `json:"toRecipients"`
+		CCRecipients      []m365Recipient    `json:"ccRecipients"`
+		BCCRecipients     []m365Recipient    `json:"bccRecipients"`
+		ReceivedDateTime  string             `json:"receivedDateTime"`
+		Body              m365Body           `json:"body"`
+		BodyPreview       string             `json:"bodyPreview"`
+		InternetMessageID string             `json:"internetMessageId"`
+		Attachments       []m365Attachment   `json:"attachments"`
+	}
+	if err := c.getJSON(ctx, path, &raw); err != nil {
+		return nil, err
+	}
+	m := &Message{
+		Source:        "m365",
+		ID:            raw.ID,
+		Folder:        "m365",
+		Subject:       raw.Subject,
+		From:          formatRecipient(raw.From),
+		To:            formatRecipients(raw.ToRecipients),
+		CC:            formatRecipients(raw.CCRecipients),
+		BCC:           formatRecipients(raw.BCCRecipients),
+		MimeMessageID: raw.InternetMessageID,
+	}
+	if t, err := time.Parse(time.RFC3339, raw.ReceivedDateTime); err == nil {
+		m.ReceivedAt = t
+	}
+	switch strings.ToLower(raw.Body.ContentType) {
+	case "html":
+		m.HTMLBody = raw.Body.Content
+		if raw.BodyPreview != "" {
+			m.TextBody = raw.BodyPreview
+		}
+	default:
+		m.TextBody = raw.Body.Content
+		if raw.BodyPreview != "" {
+			m.HTMLBody = raw.BodyPreview
+		}
+	}
+	for _, a := range raw.Attachments {
+		if a.IsInline || a.ID == "" || a.Name == "" {
+			continue
+		}
+		m.Attachments = append(m.Attachments, Attachment{
+			FileID:      a.ID,
+			FileName:    a.Name,
+			StoredName:  a.Name,
+			Size:        a.Size,
+			ContentType: a.ContentType,
+		})
+	}
+	m.HasAttachments = len(m.Attachments) > 0
+	return m, nil
+}
+
+type m365Recipient struct {
+	EmailAddress struct {
+		Name    string `json:"name"`
+		Address string `json:"address"`
+	} `json:"emailAddress"`
+}
+
+type m365Body struct {
+	ContentType string `json:"contentType"`
+	Content     string `json:"content"`
+}
+
+type m365Attachment struct {
+	ID          string `json:"id"`
+	Name        string `json:"name"`
+	ContentType string `json:"contentType"`
+	Size        int64  `json:"size"`
+	IsInline    bool   `json:"isInline"`
+	ContentID   string `json:"contentId"`
+}
+
+func formatRecipients(rs []m365Recipient) string {
+	var parts []string
+	for _, r := range rs {
+		if s := formatRecipient(r); s != "" {
+			parts = append(parts, s)
+		}
+	}
+	return strings.Join(parts, ", ")
+}
+
+func formatRecipient(r m365Recipient) string {	a := r.EmailAddress.Address
+	n := r.EmailAddress.Name
+	switch {
+	case n == "" || n == a:
+		return a
+	case a == "":
+		return n
+	default:
+		return fmt.Sprintf("%s <%s>", n, a)
+	}
+}
+
+// DownloadAttachment fetches an attachment's raw bytes via the /$value stream.
+func (c *M365Client) DownloadAttachment(ctx context.Context, mailbox, msgID, attID string) ([]byte, error) {
+	path := fmt.Sprintf("/v1.0/users/%s/messages/%s/attachments/%s/$value",
+		pathEscape(mailbox), pathEscape(msgID), pathEscape(attID))
+	tok, err := c.accessToken(ctx)
+	if err != nil {
+		return nil, err
+	}
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, c.base+path, nil)
+	if err != nil {
+		return nil, err
+	}
+	req.Header.Set("Authorization", "Bearer "+tok)
+	resp, err := c.client.Do(req)
+	if err != nil {
+		return nil, fmt.Errorf("m365 attachment: %w", err)
+	}
+	defer resp.Body.Close()
+	body, _ := io.ReadAll(io.LimitReader(resp.Body, 256<<20))
+	if resp.StatusCode != http.StatusOK {
+		return nil, fmt.Errorf("m365 attachment %s: status %d: %s", attID, resp.StatusCode, truncate(string(body), 300))
+	}
+	return body, nil
+}
+
+func (c *M365Client) getJSON(ctx context.Context, path string, out any) error {
+	tok, err := c.accessToken(ctx)
+	if err != nil {
+		return err
+	}
+	u := path
+	if !strings.HasPrefix(u, "http") {
+		u = c.base + u
+	}
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, u, nil)
+	if err != nil {
+		return err
+	}
+	req.Header.Set("Authorization", "Bearer "+tok)
+	resp, err := c.client.Do(req)
+	if err != nil {
+		return fmt.Errorf("m365 %s: %w", path, err)
+	}
+	defer resp.Body.Close()
+	body, _ := io.ReadAll(io.LimitReader(resp.Body, 16<<20))
+	if resp.StatusCode != http.StatusOK {
+		return fmt.Errorf("m365 %s: status %d: %s", path, resp.StatusCode, truncate(string(body), 300))
+	}
+	if out != nil {
+		return json.Unmarshal(body, out)
+	}
+	return nil
+}
+
+// m365Source adapts a mailbox to the Source worker-pool contract. Each mailbox
+// gets its own folder under var/mail/m365/<localpart>/ and a delta state file.
+type m365Source struct {
+	c         *M365Client
+	mailbox   string
+	localpart  string
+	stateDir   string
+	pending    string // delta link to persist on Commit()
+	hasPending bool
+}
+
+func (s *m365Source) Folder() string { return filepath.Join("m365", s.localpart) }
+
+func (s *m365Source) ListIDs(ctx context.Context, limit int, cursor string) ([]string, string, error) {
+	link, _ := os.ReadFile(filepath.Join(s.stateDir, s.localpart+".deltalink"))
+	ids, newLink, err := s.c.ListDeltaIDs(ctx, s.mailbox, strings.TrimSpace(string(link)), limit)
+	if err != nil {
+		return nil, "", err
+	}
+	// Buffer the new delta link; persist it only in Commit() after the full
+	// batch downloaded, so a failed run stays retryable without gaps.
+	if newLink != "" {
+		s.pending = newLink
+		s.hasPending = true
+	}
+	return ids, "", nil
+}
+
+// Commit persists the buffered delta link. Called by the sync runner only when
+// every listed message downloaded successfully.
+func (s *m365Source) Commit() error {
+	if !s.hasPending || s.pending == "" {
+		return nil
+	}
+	if err := os.MkdirAll(s.stateDir, 0o755); err != nil {
+		return err
+	}
+	return os.WriteFile(filepath.Join(s.stateDir, s.localpart+".deltalink"), []byte(s.pending), 0o644)
+}
+
+func (s *m365Source) Get(ctx context.Context, id string) (*Message, error) {
+	return s.c.GetMessage(ctx, s.mailbox, id)
+}
+
+func (s *m365Source) DownloadAttachment(ctx context.Context, msg *Message, att Attachment) ([]byte, error) {
+	return s.c.DownloadAttachment(ctx, s.mailbox, msg.ID, att.FileID)
+}
+
+func pathEscape(s string) string {
+	return url.PathEscape(s)
+}

bin/mail/sync/m365_test.go

@@ -0,0 +1,188 @@
+package sync
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"os"
+	"path/filepath"
+	"testing"
+)
+
+// newM365TestClient serves the Graph delta + message endpoints against a fake
+// token endpoint, so unit tests never touch the network.
+func newM365TestClient(t *testing.T, graph http.Handler) *M365Client {
+	t.Helper()
+	tok := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+		w.Write([]byte(`{"access_token":"test-token","expires_in":3600}`))
+	}))
+	gr := httptest.NewServer(graph)
+	t.Cleanup(func() {
+		tok.Close()
+		gr.Close()
+	})
+	client, err := NewM365Client(M365Credentials{Tenant: "t.onmicrosoft.com", ClientID: "c", ClientSecret: "s"})
+	if err != nil {
+		t.Fatal(err)
+	}
+	client.base = gr.URL
+	client.tokenEndpoint = tok.URL
+	return client
+}
+
+func TestM365AccessToken(t *testing.T) {
+	c := newM365TestClient(t, http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {}))
+	got, err := c.accessToken(context.Background())
+	if err != nil {
+		t.Fatalf("accessToken: %v", err)
+	}
+	if got != "test-token" {
+		t.Errorf("token = %q", got)
+	}
+	// Second call must reuse the cached token (no token request).
+	again, err := c.accessToken(context.Background())
+	if err != nil || again != "test-token" {
+		t.Fatalf("cached token: %q, %v", again, err)
+	}
+}
+
+func TestM365DeltaSkipsTombstones(t *testing.T) {
+	first := true
+	graph := http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+		if first {
+			first = false
+			w.Write([]byte(`{"value":[
+				{"id":"m1"},
+				{"id":"m2","@odata.removedReason":"deleted"},
+				{"id":"m3"}
+			],"@odata.deltaLink":"` + deltaNext + `"}`))
+			return
+		}
+		// Second call must use the stored deltaLink (points at this server).
+		if r.URL.Path != "/v1.0/delta-next" {
+			w.WriteHeader(500)
+			w.Write([]byte(`{"error":{"message":"unexpected path"}}`))
+			return
+		}
+		w.Write([]byte(`{"value":[{"id":"m4"}],"@odata.deltaLink":"` + deltaFinal + `"}`))
+	})
+	c := newM365TestClient(t, graph)
+	deltaNext = c.base + "/v1.0/delta-next"
+	deltaFinal = c.base + "/v1.0/delta-final"
+
+	ids, link, err := c.ListDeltaIDs(context.Background(), "a@x.de", "", 0)
+	if err != nil {
+		t.Fatalf("delta: %v", err)
+	}
+	if len(ids) != 2 || ids[0] != "m1" || ids[1] != "m3" {
+		t.Errorf("ids = %v", ids)
+	}
+	if link == "" {
+		t.Error("expected new deltaLink")
+	}
+	// Incremental: pass the deltaLink, get only the new id.
+	ids2, link2, err := c.ListDeltaIDs(context.Background(), "a@x.de", link, 0)
+	if err != nil {
+		t.Fatalf("delta incremental: %v", err)
+	}
+	if len(ids2) != 1 || ids2[0] != "m4" {
+		t.Errorf("ids2 = %v", ids2)
+	}
+	if link2 == "" {
+		t.Error("expected updated deltaLink")
+	}
+}
+
+func TestM365GetMessageNormalizes(t *testing.T) {
+	graph := http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+		if pathLast(r.URL.Path) == "messages" {
+			w.Write([]byte(`{"value":[{"id":"m1"}]}`))
+			return
+		}
+		w.Write([]byte(`{
+			"id":"m1",
+			"subject":"Hallo",
+			"from":{"emailAddress":{"name":"Max","address":"max@x.de"}},
+			"toRecipients":[{"emailAddress":{"address":"a@x.de"}}],
+			"receivedDateTime":"2026-08-14T08:15:00Z",
+			"body":{"contentType":"html","content":"<p>body</p>"},
+			"bodyPreview":"body",
+			"internetMessageId":"<mid@x.de>",
+			"attachments":[
+				{"id":"att1","name":"doc.pdf","contentType":"application/pdf","size":10},
+				{"id":"img1","name":"logo.png","contentType":"image/png","isInline":true}
+			]
+		}`))
+	})
+	c := newM365TestClient(t, graph)
+	m, err := c.GetMessage(context.Background(), "a@x.de", "m1")
+	if err != nil {
+		t.Fatalf("GetMessage: %v", err)
+	}
+	if m.ID != "m1" || m.Subject != "Hallo" || m.From != "Max <max@x.de>" || m.To != "a@x.de" {
+		t.Errorf("headers mismatch: %+v", m)
+	}
+	if m.HTMLBody != "<p>body</p>" {
+		t.Errorf("html = %q", m.HTMLBody)
+	}
+	if m.ReceivedAt.IsZero() {
+		t.Error("receivedAt zero")
+	}
+	if m.MimeMessageID != "<mid@x.de>" {
+		t.Errorf("mime id = %q", m.MimeMessageID)
+	}
+	if len(m.Attachments) != 1 || m.Attachments[0].FileName != "doc.pdf" || m.Attachments[0].FileID != "att1" {
+		t.Errorf("atts = %+v", m.Attachments)
+	}
+	if !m.HasAttachments {
+		t.Error("expected hasAttachments")
+	}
+}
+
+func TestM365SourceDeltaState(t *testing.T) {
+	graph := http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.Header().Set("Content-Type", "application/json")
+		w.Write([]byte(`{"value":[{"id":"m1"}],"@odata.deltaLink":"` + deltaNext + `"}`))
+	})
+	c := newM365TestClient(t, graph)
+	deltaNext = c.base + "/v1.0/delta-next"
+	stateDir := filepath.Join(t.TempDir(), ".m365")
+	s := &m365Source{c: c, mailbox: "info@x.de", localpart: "info", stateDir: stateDir}
+
+	if s.Folder() != "m365/info" {
+		t.Errorf("folder = %q", s.Folder())
+	}
+	ids, _, err := s.ListIDs(context.Background(), 0, "")
+	if err != nil {
+		t.Fatalf("ListIDs: %v", err)
+	}
+	if len(ids) != 1 || ids[0] != "m1" {
+		t.Errorf("ids = %v", ids)
+	}
+	if err := s.Commit(); err != nil {
+		t.Fatalf("Commit: %v", err)
+	}
+	data, err := os.ReadFile(filepath.Join(stateDir, "info.deltalink"))
+	if err != nil {
+		t.Fatalf("read delta state: %v", err)
+	}
+	if string(data) != deltaNext {
+		t.Errorf("delta state = %q, want %q", string(data), deltaNext)
+	}
+}
+
+func pathLast(p string) string {
+	for i := len(p) - 1; i >= 0; i-- {
+		if p[i] == '/' {
+			return p[i+1:]
+		}
+	}
+	return p
+}
+
+// deltaNext/deltaFinal are set per-test from the fake graph server URL so
+// deltaLink values always point back at the fake (never the real Graph).
+var deltaNext, deltaFinal string

bin/mail/sync/sync.go

@@ -106,6 +106,7 @@ func Retry(ctx context.Context, policy RetryPolicy, fn func() error) error {
 type SyncConfig struct {
 	OO      *OOConfig          // OnlyOffice source (optional)
 	Gmail   *GmailCredentials  // Gmail source (optional)
+	M365    *M365Credentials   // Microsoft 365 Graph source (optional)
 	Out     string             // var/mail root; default <repo>/var/mail
 	Workers int                // concurrency; default 4
 	Limit   int                // max messages per source (0 = all)
@@ -133,6 +134,14 @@ type Source interface {
 	Folder() string
 }
 
+// Committer is an optional Source capability: Commit is called after all listed
+// ids have been downloaded successfully. Sources that only advance durable state
+// on success (e.g. a Graph delta link) implement this so a killed or failed run
+// stays retryable without gaps.
+type Committer interface {
+	Commit() error
+}
+
 type ooSource struct {
 	c    *OOClient
 	page int
@@ -222,8 +231,22 @@ func Run(ctx context.Context, cfg SyncConfig) (*SyncStats, error) {
 		}
 		sources = append(sources, &gmailSource{c: gm, query: cfg.Query})
 	}
+	if cfg.M365 != nil {
+		stateDir := filepath.Join(cfg.Out, ".m365")
+		for _, mb := range cfg.M365.Users {
+			if !strings.Contains(mb, "@") {
+				return nil, fmt.Errorf("m365 user %q is not an email address", mb)
+			}
+			c, err := NewM365Client(*cfg.M365)
+			if err != nil {
+				return nil, fmt.Errorf("m365 init for %s: %w", mb, err)
+			}
+			local := strings.SplitN(mb, "@", 2)[0]
+			sources = append(sources, &m365Source{c: c, mailbox: mb, localpart: strings.ToLower(local), stateDir: stateDir})
+		}
+	}
 	if len(sources) == 0 {
-		return nil, errors.New("sync: no source configured (need OO, Gmail, or both)")
+		return nil, errors.New("sync: no source configured (need OO, Gmail, M365, or a combination)")
 	}
 
 	stats := &SyncStats{}
@@ -296,6 +319,25 @@ func Run(ctx context.Context, cfg SyncConfig) (*SyncStats, error) {
 	close(jobsCh)
 	wg.Wait()
 
+	// Only advance durable source state (e.g. delta links) when everything
+	// downloaded. A killed or failed run must be retryable without gaps.
+	if len(failures) == 0 {
+		seen := map[Source]bool{}
+		for _, j := range jobs {
+			if seen[j.src] {
+				continue
+			}
+			seen[j.src] = true
+			if c, ok := j.src.(Committer); ok {
+				if err := c.Commit(); err != nil {
+					mu.Lock()
+					failures = append(failures, j.src.Folder()+"/commit: "+err.Error())
+					mu.Unlock()
+				}
+			}
+		}
+	}
+
 	if len(failures) > 0 {
 		fmt.Fprintf(os.Stderr, "sync: %d failures:\n  %s\n", len(failures), strings.Join(failures, "\n  "))
 	}

bin/markdown/import.go

@@ -15,6 +15,7 @@ import (
 	"os"
 	"strings"
 
+	cliparse "github.com/eSlider/2dph/internal/cli"
 	"github.com/eSlider/2dph/internal/mdleaves"
 )
 
@@ -23,29 +24,13 @@ func main() {
 }
 
 func run(args []string) int {
-	jsonOut := false
+	c, err := mdleaves.ParseArgs(args)
-	files := ""
+	if err != nil {
-	root := "."
+		return cliparse.Fail(err)
-	for i := 0; i < len(args); i++ {
-		a := args[i]
-		switch {
-		case a == "--json":
-			jsonOut = true
-		case a == "--files" && i+1 < len(args):
-			i++
-			files = args[i]
-		case strings.HasPrefix(a, "--files="):
-			files = strings.TrimPrefix(a, "--files=")
-		case a == "-h" || a == "--help":
-			fmt.Fprintln(os.Stderr, "bin/markdown/import.go [dir] [--files a.md,b.md] [--json]")
-			return 0
-		case strings.HasPrefix(a, "-"):
-			fmt.Fprintln(os.Stderr, "unknown arg:", a)
-			return 2
-		default:
-			root = a
-		}
 	}
+	jsonOut := c.JSONOut
+	files := c.Files
+	root := c.Root
 
 	var paths []string
 	if files != "" {

bin/qa/stats.go

@@ -16,8 +16,8 @@ import (
 	"fmt"
 	"io"
 	"os"
-	"strings"
 
+	cliparse "github.com/eSlider/2dph/internal/cli"
 	"github.com/eSlider/2dph/internal/duckstats"
 )
 
@@ -26,23 +26,11 @@ func main() {
 }
 
 func run(args []string) int {
-	jsonl := ""
+	c, err := cliparse.ParseQAStats(args)
-	for i := 0; i < len(args); i++ {
+	if err != nil {
-		a := args[i]
+		return cliparse.Fail(err)
-		switch {
-		case a == "--jsonl" && i+1 < len(args):
-			i++
-			jsonl = args[i]
-		case strings.HasPrefix(a, "--jsonl="):
-			jsonl = strings.TrimPrefix(a, "--jsonl=")
-		case a == "-h" || a == "--help":
-			fmt.Fprintln(os.Stderr, "bin/qa/stats.go [--jsonl FILE]  # stdin = JSON [float,…]")
-			return 0
-		default:
-			fmt.Fprintln(os.Stderr, "unknown arg:", a)
-			return 2
-		}
 	}
+	jsonl := c.JSONL
 	if jsonl != "" {
 		n, err := duckstats.CountJSONL(jsonl)
 		if err != nil {

bin/reasoner/bakeoff.go

@@ -15,8 +15,8 @@ import (
 	"encoding/json"
 	"fmt"
 	"os"
-	"strings"
 
+	cliparse "github.com/eSlider/2dph/internal/cli"
 	"github.com/eSlider/2dph/internal/duckstats"
 	"github.com/eSlider/2dph/internal/reasoner"
 )
@@ -26,42 +26,13 @@ func main() {
 }
 
 func run(args []string) int {
-	base := os.Getenv("REASONER_BASE_URL")
+	c, err := reasoner.ParseArgs(args)
-	if base == "" {
+	if err != nil {
-		base = "http://127.0.0.1:11435/v1"
+		return cliparse.Fail(err)
 	}
-	model := os.Getenv("REASONER_MODEL")
+	base, model, jsonOut, device := c.Base, c.Model, c.JSONOut, c.Device
-	if model == "" {
+	client := reasoner.Client{BaseURL: base, Model: model, Device: device}
-		model = reasoner.OllamaRAM
+	rep := reasoner.Run(client)
-	}
-	jsonOut := false
-	device := "cpu"
-	for i := 0; i < len(args); i++ {
-		a := args[i]
-		switch {
-		case a == "--json":
-			jsonOut = true
-		case a == "--model" && i+1 < len(args):
-			i++
-			model = args[i]
-		case strings.HasPrefix(a, "--model="):
-			model = strings.TrimPrefix(a, "--model=")
-		case a == "--base-url" && i+1 < len(args):
-			i++
-			base = args[i]
-		case a == "--device" && i+1 < len(args):
-			i++
-			device = args[i]
-		case a == "-h" || a == "--help":
-			fmt.Fprintln(os.Stderr, "bin/reasoner/bakeoff.go [--model ID] [--base-url URL] [--device cpu] [--json]")
-			return 0
-		default:
-			fmt.Fprintln(os.Stderr, "unknown arg:", a)
-			return 2
-		}
-	}
-	c := reasoner.Client{BaseURL: base, Model: model, Device: device}
-	rep := reasoner.Run(c)
 	lat := make([]float64, 0, len(rep.Prompts))
 	for _, p := range rep.Prompts {
 		lat = append(lat, float64(p.LatencyMS))

bin/stack/lib.sh

@@ -0,0 +1,248 @@
+# bin/stack/lib.sh — compose helpers for start / start-assistant / stop / status.
+# Sourced, not executed. No secrets. No host-absolute paths.
+
+BRAIN_URL="${BRAIN_URL:-http://127.0.0.1:8630}"
+REASONER_URL="${REASONER_URL:-http://127.0.0.1:11435}"
+PICOCLAW_URL="${PICOCLAW_URL:-http://127.0.0.1:18790}"
+REASONER_MODEL="${REASONER_MODEL:-qwen3.5:9b}"
+STACK_WAIT_SECS="${STACK_WAIT_SECS:-90}"
+STACK_WAIT_INTERVAL="${STACK_WAIT_INTERVAL:-2}"
+STACK_PULL_SECS="${STACK_PULL_SECS:-600}"
+
+if [[ -z "${ROOT:-}" ]]; then
+	STACK_DIR="$(CDPATH= cd -- "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+	ROOT="$(CDPATH= cd -- "$STACK_DIR/../.." && pwd)"
+fi
+
+stack_usage() {
+	awk 'NR == 1 { next } /^#/ { sub(/^# ?/, ""); print; next } { exit }' "$1"
+}
+
+stack_die() {
+	echo "bin/stack: $*" >&2
+	return 1
+}
+
+compose() {
+	docker compose -f "$ROOT/compose.yaml" --project-directory "$ROOT" "$@"
+}
+
+http_get() {
+	local url=$1
+	local timeout=${2:-5}
+	curl -sS --max-time "$timeout" "$url" 2>/dev/null || return 1
+}
+
+health_ok() {
+	local url=$1
+	local timeout=${2:-5}
+	local body
+	body=$(http_get "$url" "$timeout") || return 1
+	printf '%s' "$body" | grep -q '"status":"ok"'
+}
+
+wait_health() {
+	local url=$1
+	local n=0
+	while ((n <= STACK_WAIT_SECS)); do
+		if health_ok "$url"; then
+			return 0
+		fi
+		n=$((n + 1))
+		if ((n <= STACK_WAIT_SECS)); then
+			sleep "$STACK_WAIT_INTERVAL"
+		fi
+	done
+	return 1
+}
+
+wait_http() {
+	local url=$1
+	local n=0
+	while ((n <= STACK_WAIT_SECS)); do
+		if http_get "$url" 5 >/dev/null; then
+			return 0
+		fi
+		n=$((n + 1))
+		if ((n <= STACK_WAIT_SECS)); then
+			sleep "$STACK_WAIT_INTERVAL"
+		fi
+	done
+	return 1
+}
+
+mcp_body() {
+	curl -sS --max-time 10 \
+		-H 'Content-Type: application/json' \
+		-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}' \
+		"$BRAIN_URL/mcp" 2>/dev/null || return 1
+}
+
+mcp_ok() {
+	local body
+	body=$(mcp_body) || return 1
+	printf '%s' "$body" | grep -Eq '"name": ?"search"' || return 1
+	printf '%s' "$body" | grep -Eq '"name": ?"get"' || return 1
+	printf '%s' "$body" | grep -Eq '"name": ?"audit"' || return 1
+	return 0
+}
+
+reasoner_tags() {
+	http_get "$REASONER_URL/api/tags" 5
+}
+
+reasoner_has_model() {
+	local body
+	body=$(reasoner_tags) || return 1
+	printf '%s' "$body" | grep -Fq "$REASONER_MODEL"
+}
+
+ensure_mcp() {
+	mcp_ok || stack_die "MCP tools/list missing search/get/audit at $BRAIN_URL/mcp"
+}
+
+ensure_brain() {
+	if health_ok "$BRAIN_URL/health"; then
+		echo "brain: reuse $BRAIN_URL" >&2
+	else
+		echo "brain: compose up" >&2
+		compose up -d brain
+		wait_health "$BRAIN_URL/health" || stack_die "brain health failed at $BRAIN_URL/health"
+	fi
+	ensure_mcp
+}
+
+pull_reasoner_model() {
+	echo "reasoner: pulling $REASONER_MODEL (CPU, may take minutes)" >&2
+	curl -sS --max-time "$STACK_PULL_SECS" \
+		-H 'Content-Type: application/json' \
+		-d "{\"name\":\"$REASONER_MODEL\"}" \
+		"$REASONER_URL/api/pull" >/dev/null
+}
+
+ensure_reasoner() {
+	if reasoner_has_model; then
+		echo "reasoner: reuse $REASONER_URL model $REASONER_MODEL" >&2
+		return 0
+	fi
+	if ! reasoner_tags >/dev/null; then
+		echo "reasoner: compose up" >&2
+		compose --profile reasoner up -d reasoner
+		wait_http "$REASONER_URL/api/tags" || stack_die "reasoner not listening at $REASONER_URL"
+	fi
+	if reasoner_has_model; then
+		return 0
+	fi
+	pull_reasoner_model
+	reasoner_has_model || stack_die "reasoner missing model $REASONER_MODEL"
+}
+
+ensure_picoclaw() {
+	echo "picoclaw: compose up --no-deps (reuse healthy :8630/:11435)" >&2
+	compose --profile picoclaw up -d --no-deps picoclaw
+	wait_health "$PICOCLAW_URL/health" || stack_die "picoclaw health failed at $PICOCLAW_URL/health"
+}
+
+mail_sync_running() {
+	compose ps --status running --services 2>/dev/null | grep -qx mail-sync
+}
+
+stack_status() {
+	local bh=down mcp=down ph=down present=false ms=down
+	health_ok "$BRAIN_URL/health" && bh=ok
+	mcp_ok && mcp=ok
+	reasoner_has_model && present=true
+	health_ok "$PICOCLAW_URL/health" && ph=ok
+	mail_sync_running && ms=ok
+	cat <<EOF
+brain:
+  url: $BRAIN_URL
+  health: $bh
+  mcp: $mcp
+reasoner:
+  url: $REASONER_URL
+  model: $REASONER_MODEL
+  present: $present
+picoclaw:
+  url: $PICOCLAW_URL
+  health: $ph
+mail_sync:
+  service: mail-sync
+  running: $ms
+EOF
+}
+
+stack_start() {
+	ensure_brain
+}
+
+stack_start_mail_sync() {
+	echo "mail-sync: compose up (ETL sync→import; index only if MAIL_SYNC_INDEX=1)" >&2
+	compose up -d mail-sync
+}
+
+stack_attach_agent() {
+	local opts=()
+	if [[ -t 0 && -t 1 ]]; then
+		opts+=(-it)
+	else
+		opts+=(-T)
+	fi
+	if [[ (! -t 0 || ! -t 1) && $# -eq 0 ]]; then
+		echo "picoclaw: no TTY. Attach with:" >&2
+		echo "  $ROOT/bin/stack/start-assistant" >&2
+		echo "  docker compose --profile picoclaw exec -it picoclaw picoclaw agent" >&2
+		return 0
+	fi
+	echo "picoclaw: agent (search → get → audit before a factual reply)" >&2
+	exec docker compose -f "$ROOT/compose.yaml" --project-directory "$ROOT" \
+		--profile picoclaw exec "${opts[@]}" picoclaw picoclaw agent "$@"
+}
+
+stack_start_assistant() {
+	local attach=1
+	local agent_args=()
+	while (($#)); do
+		case "$1" in
+		-h | --help)
+			stack_usage "$ROOT/bin/stack/start-assistant"
+			return 0
+			;;
+		--no-attach)
+			attach=0
+			shift
+			;;
+		--)
+			shift
+			agent_args+=("$@")
+			break
+			;;
+		*)
+			agent_args+=("$1")
+			shift
+			;;
+		esac
+	done
+	stack_start
+	ensure_reasoner
+	ensure_picoclaw
+	stack_status
+	if ((attach == 0)); then
+		echo "picoclaw: gateway $PICOCLAW_URL (agent not attached)" >&2
+		echo "ask the brain: $ROOT/bin/stack/start-assistant" >&2
+		echo "one-shot: $ROOT/bin/stack/start-assistant -- -m \"search the 2dph brain for LadybugDB\"" >&2
+		return 0
+	fi
+	stack_attach_agent "${agent_args[@]}"
+}
+
+stack_stop() {
+	case "${1:-}" in
+	-h | --help)
+		stack_usage "$ROOT/bin/stack/stop"
+		return 0
+		;;
+	esac
+	echo "stack: stop brain brain-mcp reasoner picoclaw mail-sync (volumes kept)" >&2
+	compose --profile picoclaw --profile reasoner stop picoclaw brain-mcp reasoner brain mail-sync
+}

bin/stack/start

@@ -0,0 +1,21 @@
+#!/usr/bin/env bash
+# bin/stack/start - bring up brain HTTP/MCP and wait until search/get/audit respond.
+#
+#   bin/stack/start
+#   bin/stack/status
+#
+# Reuses a healthy process on :8630 (host serve or compose). Does not start
+# PicoClaw. Does not rebuild Ladybug.
+set -euo pipefail
+
+STACK_DIR="$(CDPATH= cd -- "$(dirname "$0")" && pwd)"
+# shellcheck source=lib.sh
+source "$STACK_DIR/lib.sh"
+case "${1:-}" in
+-h | --help)
+	stack_usage "$0"
+	exit 0
+	;;
+esac
+stack_start "$@"
+stack_status

bin/stack/start-assistant

@@ -0,0 +1,15 @@
+#!/usr/bin/env bash
+# bin/stack/start-assistant - start + CPU reasoner + PicoClaw, then attach agent.
+#
+#   bin/stack/start-assistant
+#   bin/stack/start-assistant --no-attach
+#   bin/stack/start-assistant -- -m "search the 2dph brain for LadybugDB"
+#
+# Pulls qwen3.5:9b if missing. Gateway :18790. Agent uses MCP search → get → audit.
+# --no-attach leaves the gateway up without exec.
+set -euo pipefail
+
+STACK_DIR="$(CDPATH= cd -- "$(dirname "$0")" && pwd)"
+# shellcheck source=lib.sh
+source "$STACK_DIR/lib.sh"
+stack_start_assistant "$@"

bin/stack/start-mail-sync

@@ -0,0 +1,21 @@
+#!/usr/bin/env bash
+# bin/stack/start-mail-sync - compose up mail-sync ETL (sync → import; optional index).
+#
+#   bin/stack/start-mail-sync
+#
+# Default: onlyoffice,gmail every 300s into kb-var. Full --rebuild only if
+# MAIL_SYNC_INDEX=1 in compose/env. Secrets: ~/.config/brain/mail.env +
+# ~/.gmail-mcp (mounted). Does not start brain/picoclaw.
+set -euo pipefail
+
+STACK_DIR="$(CDPATH= cd -- "$(dirname "$0")" && pwd)"
+# shellcheck source=lib.sh
+source "$STACK_DIR/lib.sh"
+case "${1:-}" in
+-h | --help)
+	stack_usage "$0"
+	exit 0
+	;;
+esac
+stack_start_mail_sync "$@"
+stack_status

bin/stack/status

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+# bin/stack/status - YAML health for brain MCP, reasoner model, PicoClaw gateway.
+#
+#   bin/stack/status
+#   bin/stack/status | yq '.picoclaw'
+set -euo pipefail
+
+STACK_DIR="$(CDPATH= cd -- "$(dirname "$0")" && pwd)"
+# shellcheck source=lib.sh
+source "$STACK_DIR/lib.sh"
+case "${1:-}" in
+-h | --help)
+	stack_usage "$0"
+	exit 0
+	;;
+esac
+stack_status

bin/stack/stop

@@ -0,0 +1,13 @@
+#!/usr/bin/env bash
+# bin/stack/stop - stop compose brain / brain-mcp / reasoner / picoclaw / mail-sync.
+#
+#   bin/stack/stop
+#
+# Volumes kept (kb, reasoner weights, picoclaw-home). Does not kill a host
+# bin/brain/serve.go that is not a compose service.
+set -euo pipefail
+
+STACK_DIR="$(CDPATH= cd -- "$(dirname "$0")" && pwd)"
+# shellcheck source=lib.sh
+source "$STACK_DIR/lib.sh"
+stack_stop "$@"

bin/tools/contradict.py

@@ -0,0 +1,103 @@
+"""D16 contradiction adjudication (same rules as internal/facts)."""
+from __future__ import annotations
+
+from typing import Any
+
+CONF_CONFIRMED = "confirmed"
+CONF_HYPOTHESIS = "hypothesis"
+
+RULE_UNRESOLVED = "unresolved"
+RULE_TEMPORAL = "temporal_freshness"
+RULE_AUTHORITY = "authority_pairing"
+RULE_TWO_SOURCE = "two_source"
+RULE_SINGLE = "single_source"
+
+KIND_RUNTIME = "runtime"
+KIND_CONFIG = "config"
+KIND_NARRATIVE = "narrative"
+
+
+def _independent(sources: list[dict]) -> int:
+    seen: set[str] = set()
+    for i, s in enumerate(sources):
+        sid = str(s.get("id") or "") or f"{s.get('kind', '')}#{i}"
+        seen.add(sid)
+    return len(seen)
+
+
+def _fresh_n(sources: list[dict]) -> int:
+    return sum(1 for s in sources if not s.get("stale"))
+
+
+def _strong_n(sources: list[dict]) -> int:
+    return sum(1 for s in sources if s.get("kind") in (KIND_RUNTIME, KIND_CONFIG))
+
+
+def adjudicate(claim: dict[str, Any]) -> dict[str, Any]:
+    yes = list(claim.get("yes") or [])
+    no = list(claim.get("no") or [])
+    yes_n, no_n = _independent(yes), _independent(no)
+    text = str(claim.get("text") or "")
+
+    def out(conf: str, rule: str, winner: str = "") -> dict[str, Any]:
+        return {
+            "text": text,
+            "confidence": conf,
+            "confirmed": conf == CONF_CONFIRMED,
+            "rule": rule,
+            "winner": winner,
+            "yes": yes_n,
+            "no": no_n,
+        }
+
+    if yes_n < 2 or no_n < 2:
+        if yes_n >= 2:
+            return out(CONF_CONFIRMED, RULE_TWO_SOURCE, "yes")
+        if no_n >= 2:
+            return out(CONF_CONFIRMED, RULE_TWO_SOURCE, "no")
+        return out(CONF_HYPOTHESIS, RULE_SINGLE)
+    yf, nf = _fresh_n(yes), _fresh_n(no)
+    if yf >= 2 and nf < 2:
+        return out(CONF_CONFIRMED, RULE_TEMPORAL, "yes")
+    if nf >= 2 and yf < 2:
+        return out(CONF_CONFIRMED, RULE_TEMPORAL, "no")
+    ys, ns = _strong_n(yes), _strong_n(no)
+    if ys >= 2 and ns < 2:
+        return out(CONF_CONFIRMED, RULE_AUTHORITY, "yes")
+    if ns >= 2 and ys < 2:
+        return out(CONF_CONFIRMED, RULE_AUTHORITY, "no")
+    return out(CONF_HYPOTHESIS, RULE_UNRESOLVED)
+
+
+def parse_source_field(source: str) -> tuple[str, str]:
+    """Split `a x b vs c x d` into (yes, no). Empty no if no ` vs `."""
+    if " vs " not in source:
+        return source, ""
+    yes, _, no = source.partition(" vs ")
+    return yes.strip(), no.strip()
+
+
+def check_fact_row(lid: str, source: str, loc: str, how: str, conf: str) -> list[str]:
+    """Lexicon checks for one facts leaf (no Ladybug)."""
+    problems: list[str] = []
+    src = source or ""
+    if conf == CONF_CONFIRMED:
+        if " vs " in src:
+            problems.append(f"{lid}: confirmed fact cannot keep a vs-contradiction")
+        if " x " not in src:
+            problems.append(f"{lid}: needs 2-source evidence in source, got '{source}'")
+    elif conf == CONF_HYPOTHESIS:
+        yes, no = parse_source_field(src)
+        if not no or " x " not in yes or " x " not in no:
+            problems.append(
+                f"{lid}: hypothesis contradiction needs 'a x b vs c x d', got '{source}'"
+            )
+    elif conf == "partial":
+        pass
+    else:
+        problems.append(f"{lid}: unknown confidence '{conf}'")
+    if not loc:
+        problems.append(f"{lid}: missing loc (evidence pointer)")
+    if not how:
+        problems.append(f"{lid}: missing how")
+    return problems

bin/tools/kblib.py

@@ -62,7 +62,9 @@ def init_schema(conn: ladybug.Connection) -> None:
         "CREATE NODE TABLE IF NOT EXISTS Leaf ("
         " id STRING, text STRING, root STRING, confidence STRING, "
         " sha256 STRING, source STRING, source_rev STRING, observed_at STRING, "
-        " how STRING, loc STRING, type STRING, embedding FLOAT[256], "
+        " how STRING, loc STRING, type STRING, "
+        " valid_from STRING, valid_to STRING, "
+        " embedding FLOAT[256], "
         " PRIMARY KEY(id))"
     )
     conn.execute(
@@ -91,6 +93,46 @@ def init_schema(conn: ladybug.Connection) -> None:
     conn.execute(
         "CREATE REL TABLE IF NOT EXISTS AUTHORED (FROM Commit TO Person)"
     )
+    ensure_interval_columns(conn)
+
+
+def ensure_interval_columns(conn: ladybug.Connection) -> None:
+    """D24: add valid_from/valid_to on older Leaf tables (idempotent ALTER)."""
+    for col in ("valid_from", "valid_to"):
+        try:
+            conn.execute(f"ALTER TABLE Leaf ADD {col} STRING")
+        except Exception:
+            pass
+
+
+def normalize_day(s: str) -> str:
+    s = (s or "").strip()
+    if len(s) >= 10 and s[4] == "-" and s[7] == "-":
+        return s[:10]
+    return s
+
+
+def active_at(valid_from: str, valid_to: str, as_of: str) -> bool:
+    """D24: fact interval of truth. Empty ends = always; empty as_of = no filter."""
+    as_of = normalize_day(as_of)
+    if not as_of:
+        return True
+    fro = normalize_day(valid_from)
+    to = normalize_day(valid_to)
+    if fro and as_of < fro:
+        return False
+    if to and as_of > to:
+        return False
+    return True
+
+
+def filter_as_of(hits: list[dict], as_of: str) -> list[dict]:
+    if not as_of:
+        return hits
+    return [
+        h for h in hits
+        if active_at(str(h.get("valid_from") or ""), str(h.get("valid_to") or ""), as_of)
+    ]
 
 
 def leaf_id(text: str, source: str) -> str:
@@ -99,19 +141,24 @@ def leaf_id(text: str, source: str) -> str:
 
 def upsert_leaf(conn: ladybug.Connection, *, text: str, root: str, confidence: str,
                 source: str, source_rev: str, how: str, loc: str, type_: str,
-                embedding: list[float] | None) -> str:
+                embedding: list[float] | None,
+                valid_from: str = "", valid_to: str = "") -> str:
     lid = leaf_id(text, source)
     obs = time.strftime("%Y-%m-%dT%H:%M:%SZ", time.gmtime())
+    vf = normalize_day(valid_from)
+    vt = normalize_day(valid_to)
     conn.execute(
         "MERGE (l:Leaf {id:$id}) "
         "SET l.text=$text, l.root=$root, l.confidence=$confidence, "
         "    l.sha256=$sha, l.source=$source, l.source_rev=$rev, l.observed_at=$obs, "
-        "    l.how=$how, l.loc=$location, l.type=$type"
+        "    l.how=$how, l.loc=$location, l.type=$type, "
+        "    l.valid_from=$vf, l.valid_to=$vt"
         + (", l.embedding=$emb" if embedding else ""),
         parameters={
             "id": lid, "text": text, "root": root, "confidence": confidence,
             "sha": sha256_b64(text), "source": source, "rev": source_rev,
             "obs": obs, "how": how, "location": loc, "type": type_,
+            "vf": vf, "vt": vt,
             "emb": (embedding if embedding else None),
         },
     )
@@ -121,9 +168,10 @@ def upsert_leaf(conn: ladybug.Connection, *, text: str, root: str, confidence: s
 def add_leafs(conn: ladybug.Connection, leafs: list[dict]) -> list[str]:
     """Write facts+info leafs in one transaction. Safe while FTS/HNSW exist.
 
-    Each leaf dict: text, source, optional root/confidence/source_rev/how/loc/type/embedding.
+    Each leaf dict: text, source, optional root/confidence/source_rev/how/loc/type/
-    Does not delete the database file. Measured on Ladybug 0.19: MERGE of new
+    embedding/valid_from/valid_to. Does not delete the database file. Measured on
-    ids (and updates) stays FTS+HNSW queryable; DROP INDEX is the fatal path.
+    Ladybug 0.19: MERGE of new ids (and updates) stays FTS+HNSW queryable; DROP
+    INDEX is the fatal path.
     """
     if not leafs:
         return []
@@ -148,6 +196,8 @@ def add_leafs(conn: ladybug.Connection, leafs: list[dict]) -> list[str]:
                     loc=str(lf.get("loc") or lf.get("source") or ""),
                     type_=str(lf.get("type") or lf.get("type_") or "reference"),
                     embedding=lf.get("embedding"),
+                    valid_from=str(lf.get("valid_from") or ""),
+                    valid_to=str(lf.get("valid_to") or ""),
                 )
             )
         if started:
@@ -285,29 +335,39 @@ def drop_indexes(conn: ladybug.Connection) -> None:
 def query_fts(conn: ladybug.Connection, text: str, limit: int = 10) -> list[dict]:
     r = conn.execute(
         "CALL QUERY_FTS_INDEX('Leaf', 'id', $q) "
-        "RETURN node.id, node.text, node.root, score ORDER BY score DESC LIMIT $n",
+        "RETURN node.id, node.text, node.root, score, node.valid_from, node.valid_to "
+        "ORDER BY score DESC LIMIT $n",
         parameters={"q": text, "n": limit},
     )
-    return [{"id": row[0], "text": row[1], "root": row[2], "score": row[3]} for row in r.get_all()]
+    return [
+        {
+            "id": row[0], "text": row[1], "root": row[2], "score": row[3],
+            "valid_from": row[4] or "", "valid_to": row[5] or "",
+        }
+        for row in r.get_all()
+    ]
 
 
 def query_vector(conn: ladybug.Connection, embedding: list[float], limit: int = 10) -> list[dict]:
     r = conn.execute(
         "CALL QUERY_VECTOR_INDEX('Leaf', 'Leaf_vec', $q, $n) "
-        "RETURN node.id, node.text, node.root, distance ORDER BY distance LIMIT $n",
+        "RETURN node.id, node.text, node.root, distance, node.valid_from, node.valid_to "
+        "ORDER BY distance LIMIT $n",
         parameters={"q": embedding, "n": limit},
     )
     out = []
     for row in r.get_all():
-        # distance -> similarity reasonable for cosine
         score = 1.0 - row[3] if row[3] is not None else 0.0
-        out.append({"id": row[0], "text": row[1], "root": row[2], "score": score})
+        out.append({
+            "id": row[0], "text": row[1], "root": row[2], "score": score,
+            "valid_from": row[4] or "", "valid_to": row[5] or "",
+        })
     return out
 
 
 def hybrid_search(conn: ladybug.Connection, embedding: list[float], fts_hits: list[dict],
-                  limit: int = 10) -> list[dict]:
+                  limit: int = 10, as_of: str = "") -> list[dict]:
-    """Merge FTS + vector by reciprocal rank fusion."""
+    """Merge FTS + vector by reciprocal rank fusion; optional D24 as-of filter."""
     fused: dict[str, dict] = {}
     for rank, hit in enumerate(fts_hits):
         fused.setdefault(hit["id"], {**hit, "rrf": 0.0})["rrf"] = 1.0 / (60 + rank + 1)
@@ -315,8 +375,10 @@ def hybrid_search(conn: ladybug.Connection, embedding: list[float], fts_hits: li
         entry = fused.setdefault(hit["id"], {**hit, "rrf": 0.0})
         entry["rrf"] += 1.0 / (60 + rank + 1)
         entry.setdefault("score", hit.get("score", 0.0))
+        entry.setdefault("valid_from", hit.get("valid_from") or "")
+        entry.setdefault("valid_to", hit.get("valid_to") or "")
     ranked = sorted(fused.values(), key=lambda h: h.get("rrf", 0.0), reverse=True)
-    return ranked[:limit]
+    return filter_as_of(ranked, as_of)[:limit]
 
 
 def stats(conn: ladybug.Connection) -> dict:

bin/tools/test_bin_layout.py

@@ -127,6 +127,38 @@ class BinLayoutTest(unittest.TestCase):
             self.assertIn("cmdbin.ExecFile", text)
             self.assertIn(f"bin/facts/{method.removesuffix('.go')}", text)
 
+    def test_d16_adjudication_is_cgo_free(self) -> None:
+        self.assertTrue((ROOT / "internal" / "facts" / "contradict.go").is_file())
+        go = (ROOT / "internal" / "facts" / "contradict.go").read_text()
+        py = (ROOT / "bin" / "tools" / "contradict.py").read_text()
+        audit = (ROOT / "bin" / "facts" / "audit").read_text()
+        for token in ("temporal_freshness", "authority_pairing", "unresolved"):
+            self.assertIn(token, go)
+            self.assertIn(token, py)
+        self.assertIn("contradict", audit)
+        self.assertIn(" vs ", py)
+        plan = (ROOT / "PLAN.md").read_text()
+        self.assertIn("temporal_freshness", plan)
+        self.assertIn("authority_pairing", plan)
+        shebang = (ROOT / "bin" / "facts" / "audit.go").read_text()
+        self.assertIn("contradict", shebang)
+
+    def test_d23_flaggy_cli(self) -> None:
+        self.assertTrue((ROOT / "internal" / "cli" / "cli.go").is_file())
+        self.assertIn("github.com/integrii/flaggy", (ROOT / "go.mod").read_text())
+        plan = (ROOT / "PLAN.md").read_text()
+        self.assertIn("D23", plan)
+        self.assertIn("flaggy", plan)
+        complete = (ROOT / "bin" / "cli" / "complete.go").read_text()
+        first = complete.splitlines()[0]
+        self.assertTrue(first.startswith("//usr/bin/env go run"), first)
+        self.assertIn("complete.go bash", complete)
+        self.assertIn("brain-search", complete)
+        chats_import = (ROOT / "internal" / "chats" / "import.go").read_text()
+        self.assertNotIn("flag.NewFlagSet", chats_import)
+        args = (ROOT / "internal" / "brain" / "rank" / "args.go").read_text()
+        self.assertIn("internal/cli", args)
+
     def test_mail_import_is_shebang_not_brain_write(self) -> None:
         self._assert_shebang("bin/mail/import.go")
         index_mail = (ROOT / "bin" / "mail" / "index_mail").read_text()

bin/tools/test_contradict.py

@@ -0,0 +1,104 @@
+import os
+import sys
+import unittest
+
+sys.path.insert(0, os.path.dirname(__file__))
+
+from contradict import (  # noqa: E402
+    RULE_AUTHORITY,
+    RULE_SINGLE,
+    RULE_TEMPORAL,
+    RULE_TWO_SOURCE,
+    RULE_UNRESOLVED,
+    adjudicate,
+    check_fact_row,
+    parse_source_field,
+)
+
+
+def src(i, kind, stale=False):
+    return {"id": i, "kind": kind, "stale": stale}
+
+
+class TestContradict(unittest.TestCase):
+    def test_two_vs_two_stays_hypothesis(self):
+        r = adjudicate({
+            "text": "svc listens on 443",
+            "yes": [src("docker-ps", "runtime"), src("compose", "config")],
+            "no": [src("docker-old", "runtime"), src("compose-old", "config")],
+        })
+        self.assertFalse(r["confirmed"])
+        self.assertEqual(r["rule"], RULE_UNRESOLVED)
+        self.assertEqual(r["winner"], "")
+
+    def test_temporal_freshness(self):
+        r = adjudicate({
+            "text": "svc listens on 443",
+            "yes": [src("docker-ps", "runtime"), src("compose", "config")],
+            "no": [src("old-readme", "narrative", True), src("old-wiki", "narrative", True)],
+        })
+        self.assertTrue(r["confirmed"])
+        self.assertEqual(r["rule"], RULE_TEMPORAL)
+        self.assertEqual(r["winner"], "yes")
+
+    def test_authority_pairing(self):
+        r = adjudicate({
+            "text": "svc listens on 443",
+            "yes": [src("docker-ps", "runtime"), src("compose", "config")],
+            "no": [src("readme", "narrative"), src("wiki", "narrative")],
+        })
+        self.assertTrue(r["confirmed"])
+        self.assertEqual(r["rule"], RULE_AUTHORITY)
+        self.assertEqual(r["winner"], "yes")
+
+    def test_two_source_and_single(self):
+        two = adjudicate({
+            "text": "arc-1 runs Matrix",
+            "yes": [src("compose", "config"), src("docker-ps", "runtime")],
+        })
+        self.assertTrue(two["confirmed"])
+        self.assertEqual(two["rule"], RULE_TWO_SOURCE)
+        one = adjudicate({"text": "maybe", "yes": [src("readme", "narrative")]})
+        self.assertFalse(one["confirmed"])
+        self.assertEqual(one["rule"], RULE_SINGLE)
+
+    def test_parse_source_field(self):
+        yes, no = parse_source_field("docker ps x compose.yml vs old.md x wiki.md")
+        self.assertIn(" x ", yes)
+        self.assertIn(" x ", no)
+
+    def test_check_fact_row_allows_hypothesis_vs(self):
+        p = check_fact_row(
+            "L1", "a.md x b.md vs c.md x d.md", "var/", "audit", "hypothesis",
+        )
+        self.assertEqual(p, [])
+        p = check_fact_row("L2", "a.md x b.md", "var/", "audit", "confirmed")
+        self.assertEqual(p, [])
+        p = check_fact_row("L3", "a.md x b.md vs c.md x d.md", "var/", "audit", "confirmed")
+        self.assertTrue(any("vs-contradiction" in x for x in p))
+        p = check_fact_row("L4", "only-one.md", "var/", "audit", "hypothesis")
+        self.assertTrue(any("a x b vs" in x for x in p))
+
+    def test_audit_contradict_cli_unresolved(self):
+        import json
+        import subprocess
+        from pathlib import Path
+        root = Path(__file__).resolve().parents[2]
+        payload = json.dumps({
+            "text": "svc 443",
+            "yes": [src("a", "runtime"), src("b", "config")],
+            "no": [src("c", "runtime"), src("d", "config")],
+        })
+        proc = subprocess.run(
+            [sys.executable, str(root / "bin" / "facts" / "audit"), "contradict", "--json"],
+            input=payload, capture_output=True, text=True, check=False,
+        )
+        self.assertEqual(proc.returncode, 0, proc.stderr)
+        out = json.loads(proc.stdout)
+        self.assertTrue(out["ok"])
+        self.assertEqual(out["contradictions"][0]["rule"], RULE_UNRESOLVED)
+        self.assertFalse(out["contradictions"][0]["confirmed"])
+
+
+if __name__ == "__main__":
+    unittest.main()

bin/tools/test_kblib.py

@@ -194,6 +194,38 @@ class KblibTest(unittest.TestCase):
         self.assertEqual(person["name"], "Ada Lovelace")
         self.assertEqual(person["depth"], 3)
 
+    def test_as_of_keeps_x_drops_y(self) -> None:
+        """OQ5/#36: as of 2025-01-01 → works-at-X, not works-at-Y."""
+        kblib.upsert_leaf(
+            self.conn, text="Andrey works at X", root="facts",
+            confidence="confirmed", source="crm.md x contract.md",
+            source_rev="r1", how="test", loc="/tmp", type_="fact",
+            embedding=make_emb(0.5),
+            valid_from="2024-03-01", valid_to="2025-07-15",
+        )
+        kblib.upsert_leaf(
+            self.conn, text="Andrey works at Y", root="facts",
+            confidence="confirmed", source="offer.md x payroll.md",
+            source_rev="r1", how="test", loc="/tmp", type_="fact",
+            embedding=make_emb(0.6),
+            valid_from="2025-07-16", valid_to="",
+        )
+        kblib.ensure_indexes(self.conn)
+        hits = kblib.query_fts(self.conn, "Andrey works", 10)
+        kept = kblib.filter_as_of(hits, "2025-01-01")
+        texts = [h["text"] for h in kept]
+        self.assertTrue(any("works at X" in t for t in texts), texts)
+        self.assertFalse(any("works at Y" in t for t in texts), texts)
+        later = kblib.filter_as_of(hits, "2025-08-01")
+        later_texts = [h["text"] for h in later]
+        self.assertTrue(any("works at Y" in t for t in later_texts), later_texts)
+        self.assertFalse(any("works at X" in t for t in later_texts), later_texts)
+
+    def test_active_at_pure(self) -> None:
+        self.assertTrue(kblib.active_at("2024-03-01", "2025-07-15", "2025-01-01"))
+        self.assertFalse(kblib.active_at("2025-07-16", "", "2025-01-01"))
+        self.assertTrue(kblib.active_at("", "", "2025-01-01"))
+
 
 if __name__ == "__main__":
     unittest.main()

bin/tools/test_published_docs.py

@@ -186,3 +186,21 @@ class PublishedDocsTest(unittest.TestCase):
         self.assertIn("epic #16", index)
         agents = (ROOT / "AGENTS.md").read_text()
         self.assertIn("roadmap.md", agents)
+
+    def test_oq5_intervals_are_not_d16_stale(self) -> None:
+        plan = (ROOT / "PLAN.md").read_text()
+        design = (ROOT / "docs" / "design.md").read_text()
+        road = (ROOT / "docs" / "roadmap.md").read_text()
+        contradict = (ROOT / "internal" / "facts" / "contradict.go").read_text()
+        interval = (ROOT / "internal" / "facts" / "interval.go").read_text()
+        self.assertIn("OQ5", plan)
+        self.assertIn("D24", plan)
+        self.assertIn("issues/36", plan)
+        self.assertIn("valid_from", plan)
+        self.assertIn("--as-of", design)
+        self.assertIn("issues/36", road)
+        self.assertIn("**in**", road[road.index("OQ5"):road.index("OQ5") + 80])
+        self.assertIn("temporal_freshness", contradict)
+        self.assertNotIn("valid_from", contradict)
+        self.assertIn("ActiveAt", interval)
+        self.assertIn("NormalizeDay", interval)

bin/tools/test_stack.py

@@ -0,0 +1,228 @@
+"""bin/stack/{start,start-assistant,stop,status} — offline contract + fake PATH."""
+from __future__ import annotations
+
+import os
+import stat
+import subprocess
+import tempfile
+import unittest
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parents[2]
+METHODS = ("start", "start-assistant", "start-mail-sync", "stop", "status")
+
+
+class StackLayoutTest(unittest.TestCase):
+    def test_methods_are_bash_with_usage_comment(self) -> None:
+        lib = ROOT / "bin" / "stack" / "lib.sh"
+        self.assertTrue(lib.is_file(), "missing bin/stack/lib.sh")
+        for name in METHODS:
+            p = ROOT / "bin" / "stack" / name
+            self.assertTrue(p.is_file(), f"missing bin/stack/{name}")
+            self.assertTrue(os.access(p, os.X_OK), f"bin/stack/{name} must be executable")
+            lines = p.read_text().splitlines()
+            self.assertEqual(lines[0], "#!/usr/bin/env bash", name)
+            self.assertTrue(lines[1].startswith("# bin/stack/"), name)
+            text = "\n".join(lines)
+            self.assertIn("lib.sh", text, name)
+            self.assertNotIn("/mnt/", text, name)
+            self.assertNotIn("/home/", text, name)
+
+    def test_lib_has_no_host_paths_or_secrets(self) -> None:
+        lib = (ROOT / "bin" / "stack" / "lib.sh").read_text()
+        self.assertIn("stack_start", lib)
+        self.assertIn("stack_start_assistant", lib)
+        self.assertIn("stack_start_mail_sync", lib)
+        self.assertIn("stack_stop", lib)
+        self.assertIn("stack_status", lib)
+        self.assertIn("mail-sync", lib)
+        self.assertIn("qwen3.5:9b", lib)
+        self.assertIn("picoclaw agent", lib)
+        self.assertIn("--no-deps", lib)
+        self.assertIn("tools/list", lib)
+        self.assertNotIn("/mnt/", lib)
+        self.assertNotIn("/home/", lib)
+        self.assertNotIn("password", lib.lower())
+        self.assertNotIn("GITEA_TOKEN", lib)
+
+    def test_start_does_not_launch_picoclaw(self) -> None:
+        start = (ROOT / "bin" / "stack" / "start").read_text()
+        self.assertIn("stack_start", start)
+        self.assertNotIn("stack_start_assistant", start)
+        self.assertNotIn("picoclaw agent", start)
+
+    def test_start_mail_sync_is_etl_only(self) -> None:
+        src = (ROOT / "bin" / "stack" / "start-mail-sync").read_text()
+        self.assertIn("stack_start_mail_sync", src)
+        self.assertNotIn("stack_start_assistant", src)
+        ep = (ROOT / "bin" / "docker-entrypoint").read_text()
+        self.assertIn('MAIL_SYNC_SRC:=onlyoffice,gmail', ep)
+        self.assertIn('MAIL_SYNC_INDEX:=0', ep)
+        self.assertIn('interval="${1:-300}"', ep)
+        self.assertIn('MAIL_SYNC_INDEX" = "1"', ep)
+
+    def test_start_assistant_attaches_agent(self) -> None:
+        src = (ROOT / "bin" / "stack" / "start-assistant").read_text()
+        self.assertIn("stack_start_assistant", src)
+        self.assertIn("--no-attach", src)
+
+    def test_stop_does_not_down_volumes(self) -> None:
+        lib = (ROOT / "bin" / "stack" / "lib.sh").read_text()
+        self.assertIn(" compose ", lib)
+        self.assertRegex(lib, r"\bstop\b")
+        self.assertNotIn(" compose down", lib)
+        self.assertNotIn("compose down", lib)
+
+    def test_docs_name_stack_commands(self) -> None:
+        runbook = (ROOT / "docs" / "runbook.md").read_text()
+        pico = (ROOT / "docs" / "picoclaw.md").read_text()
+        agents = (ROOT / "AGENTS.md").read_text()
+        for text in (runbook, pico, agents):
+            self.assertIn("bin/stack/start", text)
+            self.assertIn("bin/stack/start-assistant", text)
+            self.assertIn("bin/stack/status", text)
+            self.assertIn("bin/stack/stop", text)
+
+
+    def test_help_prints_comments_not_source(self) -> None:
+        r = subprocess.run(
+            [str(ROOT / "bin" / "stack" / "start"), "--help"],
+            cwd=str(ROOT),
+            capture_output=True,
+            text=True,
+            check=False,
+        )
+        self.assertEqual(r.returncode, 0, r.stderr)
+        self.assertIn("bin/stack/start", r.stdout)
+        self.assertNotIn("set -euo pipefail", r.stdout)
+        self.assertNotIn("source ", r.stdout)
+
+
+class StackFakePathTest(unittest.TestCase):
+    def _fake_bin(self, tmp: Path, *, health_ok: bool) -> Path:
+        bindir = tmp / "bin"
+        bindir.mkdir()
+        curl = bindir / "curl"
+        docker = bindir / "docker"
+        log = tmp / "docker.log"
+        curl.write_text(
+            f"""#!/usr/bin/env bash
+url=""
+for a in "$@"; do
+  case "$a" in http*) url=$a ;;
+  esac
+done
+if [ "{int(health_ok)}" = "0" ] && [[ "$url" == */health ]]; then
+  echo '{{"status":"down"}}'
+  exit 7
+fi
+case "$url" in
+  */mcp)
+    echo '{{"jsonrpc":"2.0","id":1,"result":{{"tools":[{{"name":"search"}},{{"name":"get"}},{{"name":"audit"}}]}}}}'
+    ;;
+  */api/tags)
+    echo '{{"models":[{{"name":"qwen3.5:9b"}}]}}'
+    ;;
+  */api/pull)
+    echo '{{"status":"success"}}'
+    ;;
+  *)
+    echo '{{"status":"ok"}}'
+    ;;
+esac
+"""
+        )
+        docker.write_text(
+            f"""#!/usr/bin/env bash
+echo "$*" >> "{log}"
+exit 0
+"""
+        )
+        curl.chmod(curl.stat().st_mode | stat.S_IEXEC)
+        docker.chmod(docker.stat().st_mode | stat.S_IEXEC)
+        return bindir
+
+    def _env(self, bindir: Path) -> dict[str, str]:
+        env = os.environ.copy()
+        env["PATH"] = f"{bindir}:{env.get('PATH', '')}"
+        env["STACK_WAIT_SECS"] = "1"
+        env["STACK_WAIT_INTERVAL"] = "0"
+        return env
+
+    def test_start_skips_compose_when_brain_healthy(self) -> None:
+        with tempfile.TemporaryDirectory() as raw:
+            tmp = Path(raw)
+            bindir = self._fake_bin(tmp, health_ok=True)
+            log = tmp / "docker.log"
+            r = subprocess.run(
+                [str(ROOT / "bin" / "stack" / "start")],
+                cwd=str(ROOT),
+                env=self._env(bindir),
+                capture_output=True,
+                text=True,
+                check=False,
+            )
+            self.assertEqual(r.returncode, 0, r.stderr)
+            if log.exists():
+                logged = log.read_text()
+                self.assertNotIn("up -d", logged, "healthy brain must not docker compose up")
+                self.assertIn("ps", logged)  # status probes mail-sync via compose ps
+
+    def test_start_ups_brain_when_unhealthy(self) -> None:
+        with tempfile.TemporaryDirectory() as raw:
+            tmp = Path(raw)
+            bindir = self._fake_bin(tmp, health_ok=False)
+            log = tmp / "docker.log"
+            r = subprocess.run(
+                [str(ROOT / "bin" / "stack" / "start")],
+                cwd=str(ROOT),
+                env=self._env(bindir),
+                capture_output=True,
+                text=True,
+                check=False,
+            )
+            self.assertNotEqual(r.returncode, 0, "unhealthy brain without recovering compose must fail")
+            self.assertTrue(log.exists(), r.stderr)
+            logged = log.read_text()
+            self.assertIn("up -d", logged)
+            self.assertIn("brain", logged)
+            self.assertNotIn("picoclaw", logged)
+
+    def test_stop_stops_named_services(self) -> None:
+        with tempfile.TemporaryDirectory() as raw:
+            tmp = Path(raw)
+            bindir = self._fake_bin(tmp, health_ok=True)
+            log = tmp / "docker.log"
+            r = subprocess.run(
+                [str(ROOT / "bin" / "stack" / "stop")],
+                cwd=str(ROOT),
+                env=self._env(bindir),
+                capture_output=True,
+                text=True,
+                check=False,
+            )
+            self.assertEqual(r.returncode, 0, r.stderr)
+            logged = log.read_text()
+            self.assertIn("stop", logged)
+            self.assertNotIn(" down", logged)
+            for svc in ("brain", "brain-mcp", "reasoner", "picoclaw", "mail-sync"):
+                self.assertIn(svc, logged)
+
+    def test_start_assistant_no_attach_starts_picoclaw(self) -> None:
+        with tempfile.TemporaryDirectory() as raw:
+            tmp = Path(raw)
+            bindir = self._fake_bin(tmp, health_ok=True)
+            log = tmp / "docker.log"
+            r = subprocess.run(
+                [str(ROOT / "bin" / "stack" / "start-assistant"), "--no-attach"],
+                cwd=str(ROOT),
+                env=self._env(bindir),
+                capture_output=True,
+                text=True,
+                check=False,
+            )
+            self.assertEqual(r.returncode, 0, r.stderr + r.stdout)
+            logged = log.read_text() if log.exists() else ""
+            self.assertIn("picoclaw", logged)
+            self.assertIn("--no-deps", logged)
+            self.assertNotIn("picoclaw agent", logged)

bin/web/search.go

@@ -16,9 +16,9 @@ import (
 	"fmt"
 	"net/http"
 	"os"
-	"strconv"
 	"time"
 
+	"github.com/eSlider/2dph/internal/cli"
 	"github.com/eSlider/2dph/internal/websearch"
 	"golang.org/x/sys/unix"
 )
@@ -28,77 +28,15 @@ func main() {
 }
 
 func run(args []string) int {
-	var (
+	c, err := websearch.ParseArgs(args)
-		query, site, lang, fresh, category, engines string
+	if err != nil {
-		limit                                       = websearch.DefaultLimit
+		return cli.Fail(err)
-		jsonOut, refresh, force                     bool
-		ttl                                         = float64(websearch.CacheTTL)
-		timeout                                     = 25
-	)
-	i := 0
-	for i < len(args) {
-		a := args[i]
-		switch {
-		case a == "--json":
-			jsonOut = true
-		case a == "--refresh":
-			refresh = true
-		case a == "--force":
-			force = true
-		case (a == "-n" || a == "--limit") && i+1 < len(args):
-			i++
-			n, err := strconv.Atoi(args[i])
-			if err != nil || n < 0 {
-				fmt.Fprintln(os.Stderr, "web/search: --limit must be a non-negative integer")
-				return 2
-			}
-			limit = n
-		case a == "--site" && i+1 < len(args):
-			i++
-			site = args[i]
-		case a == "--lang" && i+1 < len(args):
-			i++
-			lang = args[i]
-		case a == "--fresh" && i+1 < len(args):
-			i++
-			fresh = args[i]
-		case a == "--category" && i+1 < len(args):
-			i++
-			category = args[i]
-		case a == "--engines" && i+1 < len(args):
-			i++
-			engines = args[i]
-		case a == "--ttl" && i+1 < len(args):
-			i++
-			v, err := strconv.ParseFloat(args[i], 64)
-			if err != nil {
-				fmt.Fprintln(os.Stderr, "web/search: --ttl must be a number")
-				return 2
-			}
-			ttl = v
-		case a == "--timeout" && i+1 < len(args):
-			i++
-			n, err := strconv.Atoi(args[i])
-			if err != nil || n <= 0 {
-				fmt.Fprintln(os.Stderr, "web/search: --timeout must be a positive integer")
-				return 2
-			}
-			timeout = n
-		case a == "-h" || a == "--help":
-			fmt.Fprintln(os.Stderr, `usage: bin/web/search.go QUERY [--json] [-n N] [--site HOST] [--lang LANG] [--fresh day|week|month|year] [--category CAT] [--engines LIST] [--refresh] [--force]`)
-			return 0
-		case len(a) > 0 && a[0] != '-' && query == "":
-			query = a
-		default:
-			fmt.Fprintf(os.Stderr, "web/search: unknown flag %s\n", a)
-			return 2
-		}
-		i++
-	}
-	if query == "" {
-		fmt.Fprintln(os.Stderr, "web/search: query required")
-		return 2
 	}
+	query, site, lang, fresh, category, engines := c.Query, c.Site, c.Lang, c.Fresh, c.Category, c.Engines
+	limit := c.Limit
+	jsonOut, refresh, force := c.JSONOut, c.Refresh, c.Force
+	ttl := c.TTL
+	timeout := c.Timeout
 	if site != "" {
 		query = "site:" + site + " " + query
 	}

compose.yaml

@@ -1,5 +1,6 @@
 # 2dph — docker composition
 #
+#   bin/stack/start / start-assistant / status / stop
 #   docker compose up -d brain                  # API (Zig CGO serve)
 #   docker compose --profile index run --rm index   # Python rebuild
 #   docker compose --profile picoclaw up -d     # brain-mcp + CPU reasoner + PicoClaw gateway
@@ -91,6 +92,38 @@ services:
     tmpfs:
       - /tmp
 
+  # mail-sync ETL: sync → import every 300s into shared var. Full brain rebuild
+  # only when MAIL_SYNC_INDEX=1 (expensive on large mail corpora). Default
+  # sources: onlyoffice,gmail. For M365 set MAIL_SYNC_SRC=m365 and put
+  # M365_TENANT/CLIENT_ID/CLIENT_SECRET/USERS in ~/.config/brain/mail.env
+  # (or m365.env + MAIL_SYNC_ENV). Gmail OAuth: mount ~/.gmail-mcp.
+  #   docker compose up -d mail-sync
+  #   bin/stack/start-mail-sync
+  mail-sync:
+    image: ghcr.io/eslider/2dph:index
+    build:
+      context: .
+      dockerfile: Dockerfile
+      target: index
+    environment:
+      HF_HOME: /data/hf
+      KB_PY: python3
+      MAIL_SYNC_SRC: onlyoffice,gmail
+      MAIL_SYNC_ENV: /secret/mail.env
+      MAIL_SYNC_INDEX: "0"
+      HOME: /home/2dph
+    volumes:
+      - kb-model:/data/hf
+      - kb-var:/app/var
+      - ~/.config/brain:/secret:ro
+      - ~/.gmail-mcp:/home/2dph/.gmail-mcp:ro
+    command: ["mail-sync", "300"]
+    read_only: true
+    tmpfs:
+      - /tmp
+    restart: unless-stopped
+    stop_grace_period: 20s
+
   # Optional local SearXNG (D3). Skip if BRAIN_SEARCH_URL already points at a
   # live instance — do not run a second copy on that host.
   #   SEARXNG_SECRET=$(openssl rand -hex 32) docker compose --profile searxng up -d

docs/README.md

@@ -18,9 +18,9 @@ Evidence-first knowledge graph. Facts need proof or they are
 | tutorial / howto | [runbook](runbook.md) — run anywhere (uv, Go, Docker) |
 | explanation | [design](design.md) — two roots, deduction, D17/D20/D18 |
 | explanation | [roadmap](roadmap.md) — gap to v1 (epic #16) |
-| howto | [picoclaw](picoclaw.md) — MCP agent profile |
+| howto | [picoclaw](picoclaw.md) — MCP agent (`bin/stack/start-assistant`) |
 | howto | [reasoner](reasoner.md) — CPU bake-off (D18) |
-| reference | [PLAN.md](../PLAN.md) — decisions D1–D22 |
+| reference | [PLAN.md](../PLAN.md) — decisions D1–D24 |
 
 Decisions the public face must name: **D3** SearXNG compose, **D6** Go service /
 Python write sidecar, **D14** `bin/{subject}/{method}.go`, **D15** Gitea origin,

docs/design.md

@@ -38,6 +38,10 @@ bin/brain/search.go "question"
 from each hit (1=File, 2=Commit, 3=Person). Rebuild writes FROM_FILE;
 git import writes HAS_VERSION/AUTHORED ([#17](https://git.produktor.io/eSlider/2dph/issues/17)).
 
+Go CLIs parse with **flaggy** via `internal/cli` (D23). Flags may appear
+after positionals (`search q --hop 1`). Completions:
+`source <(./bin/cli/complete.go bash)`.
+
 ## Who / What / How / Where / When + evidence
 
 Every assertion edge carries:
@@ -64,6 +68,15 @@ binary); conversion prints leafs, brain write is `bin/brain/index.go`.
 `bin/facts/audit stale` flags leafs whose observed revision is behind the
 corpus HEAD.
 
+Fact **interval of truth** (D24 / OQ5): leaf props `valid_from` /
+`valid_to` (YYYY-MM-DD, inclusive; empty end = open; both empty = legacy
+always-active). `bin/brain/search.go --as-of YYYY-MM-DD` and MCP/HTTP
+`as_of` keep hits whose interval covers that day. This is not D16
+`temporal_freshness` (source freshness vs HEAD). [#36](https://git.produktor.io/eSlider/2dph/issues/36).
+Existing `kb.lbug` without the columns: open/search/add runs an idempotent
+`ALTER TABLE Leaf ADD …` (no full rebuild required). Fresh `--rebuild` still
+creates them in `CREATE NODE TABLE`.
+
 ## Sources (auto-pairing)
 
 - A: runtime state — `docker ps` (container running), ports actually bound
@@ -71,7 +84,9 @@ corpus HEAD.
 - C: narrative — READMEs, AGENTS.md, docs
 
 Confirmed = A×B or B×C agreement. Single source = hypothesis + `(not confirmed)`.
-Conflicting pairings (≥2 yes vs ≥2 no) = hypothesis (OQ1 → v2 resolution).
+Conflicting pairings (≥2 yes vs ≥2 no) stay hypothesis until
+`temporal_freshness` or `authority_pairing` fires (`bin/facts/audit contradict`,
+[#29](https://git.produktor.io/eSlider/2dph/issues/29)).
 
 ## Read path
 

docs/picoclaw.md

@@ -6,6 +6,18 @@ CPU reasoner. Default agent model is `qwen3.5:9b` (RAM path, D18). Weights stay
 in the reasoner volume, not in the 2dph image.
 No secrets in git: Ollama needs no key; MCP is local HTTP.
 
+```bash
+bin/stack/start-assistant
+bin/stack/start-assistant --no-attach
+bin/stack/start-assistant -- -m "search the 2dph brain for LadybugDB"
+bin/stack/status
+bin/stack/stop
+```
+
+`start-assistant` reuses a healthy brain on `:8630`, starts the CPU reasoner,
+pulls `qwen3.5:9b` if missing, brings up the gateway with `--no-deps picoclaw`,
+then `picoclaw agent` (MCP `search` → `get` → `audit`). Gateway-only Compose:
+
 ```bash
 docker compose --profile picoclaw up -d
 # already serving :8630 / :11435:

docs/roadmap.md

@@ -39,11 +39,14 @@ Epic [#16](https://git.produktor.io/eSlider/2dph/issues/16) closed.
 [#6](https://git.produktor.io/eSlider/2dph/issues/6) OCR — **in**.
 [#30](https://git.produktor.io/eSlider/2dph/issues/30) OQ3 duckdb-go — **in**.
 [#29](https://git.produktor.io/eSlider/2dph/issues/29) OQ1 contradiction
-resolution.
+resolution — **in** (`temporal_freshness`, `authority_pairing`).
+[#34](https://git.produktor.io/eSlider/2dph/issues/34) D23 flaggy CLI — **in**.
+[#36](https://git.produktor.io/eSlider/2dph/issues/36) OQ5/D24 fact intervals /
+as-of — **in** (not D16 `temporal_freshness`).
 
 ## Blockers
 
-None for epic #16 (closed). Remaining v2: OQ1, OQ4.
+None for epic #16 (closed). Remaining v2: OQ4 (deferred).
 
 ```
 question
@@ -58,8 +61,10 @@ question
 
 ## Not v1
 
-OQ1 contradiction resolution, OQ4 YAML-first leafs.
+OQ4 YAML-first leafs.
-OCR (OQ2) and duckdb-go (OQ3/D22) are in.
+OQ5/D24 fact intervals + as-of — **in**
+([#36](https://git.produktor.io/eSlider/2dph/issues/36)).
+OCR (OQ2), duckdb-go (OQ3/D22), and D16 adjudication (OQ1) are in.
 
 ## Close epic #16 when
 

docs/runbook.md

@@ -53,11 +53,15 @@ bin/brain/add.go --text "arc-1 runs Matrix" --root facts --source "compose.yml x
 bin/brain/index.go --rebuild --with-facts --with-chats
 bin/brain/search.go "LadybugDB vector index"     # facts → info → web (D17)
 bin/brain/search.go "upstream flag" --no-web
+bin/brain/search.go "who works where" --as-of 2025-01-01  # D24 intervals
+source <(./bin/cli/complete.go bash)            # D23 flaggy complete
 bin/brain/get.go <id> --body
 bin/brain/stats.go
 ```
 
-`--hop N` walks File → Commit → Person from each hit. Empty web results are `throttled`, not absence.
+`--hop N` walks File → Commit → Person from each hit. `--as-of YYYY-MM-DD`
+keeps leafs whose `valid_from`/`valid_to` cover that day (empty interval =
+legacy always-on). Empty web results are `throttled`, not absence.
 Gap to v1: [roadmap](roadmap.md) / [epic #16](https://git.produktor.io/eSlider/2dph/issues/16).
 
 Ladybug 0.19: never `DROP INDEX` FTS/VECTOR (ghost catalog). Fresh indexes =
@@ -65,13 +69,26 @@ delete `var/kb.lbug` then `--rebuild`.
 
 ## HTTP / MCP
 
+```bash
+bin/stack/start                 # brain :8630, wait until MCP search/get/audit
+bin/stack/status                # YAML: brain / reasoner / picoclaw / mail_sync
+bin/stack/start-mail-sync       # compose ETL: OO+Gmail sync→import (300s; no auto-rebuild)
+bin/stack/start-assistant       # + qwen3.5:9b + PicoClaw agent (ask the brain)
+bin/stack/start-assistant --no-attach
+bin/stack/stop                  # compose stop; volumes kept
+```
+
+Same Compose services by hand:
+
 ```bash
 docker compose up -d brain                       # :8630 Zig CGO serve
+docker compose up -d mail-sync                   # ETL loop into kb-var
 docker compose --profile index run --rm index    # rebuild
 docker compose --profile picoclaw up brain-mcp   # MCP 127.0.0.1:8630
 ```
 
 `GET /openapi.json`, `POST /mcp`. Agent tool order: `search` → `get` → `audit`.
+See [picoclaw.md](picoclaw.md).
 
 ## Reasoner (optional, D18)
 

go.mod

@@ -9,6 +9,7 @@ require (
 	github.com/daulet/tokenizers v1.27.0
 	github.com/duckdb/duckdb-go/v2 v2.10505.0
 	github.com/go-git/go-git/v5 v5.19.2
+	github.com/integrii/flaggy v1.8.0
 	golang.org/x/sys v0.47.0
 	golang.org/x/text v0.40.0
 	modernc.org/sqlite v1.56.0

go.sum

@@ -77,6 +77,8 @@ github.com/google/uuid v1.6.0 h1:NIvaJDMOsjHA8n1jAhLSgzrAzy1Hgr+hNrb57e+94F0=
 github.com/google/uuid v1.6.0/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
 github.com/hashicorp/golang-lru/v2 v2.0.7 h1:a+bsQ5rvGLjzHuww6tVxozPZFVghXaHOwFs4luLUK2k=
 github.com/hashicorp/golang-lru/v2 v2.0.7/go.mod h1:QeFd9opnmA6QUJc5vARoKUSoFhyfM2/ZepoAG6RGpeM=
+github.com/integrii/flaggy v1.8.0 h1:tC1qWwg4fhF2Qdaj+MpPK04cxlOSq0+HoMZqAW6Arao=
+github.com/integrii/flaggy v1.8.0/go.mod h1:QS4c80m87SXG0pmVUT/Lx2RY5EbkLvLp7IKBD2jwcFA=
 github.com/jbenet/go-context v0.0.0-20150711004518-d14ea06fba99 h1:BQSFePA1RWJOlocH6Fxy8MmwDt+yVQYULKfN0RoTN8A=
 github.com/jbenet/go-context v0.0.0-20150711004518-d14ea06fba99/go.mod h1:1lJo3i6rXxKeerYnT8Nvf0QmHCRC1n8sfWVwXF2Frvo=
 github.com/kevinburke/ssh_config v1.2.0 h1:x584FjTGwHzMwvHx18PXxbBVzfnxogHaAReU4gf13a4=

internal/brain/db.go

@@ -85,9 +85,18 @@ func openWithSandbox(epsv string) error {
 		closeBrain()
 		return fmt.Errorf("LOAD EXTENSION VECTOR: %w", err)
 	}
+	migrateIntervalColumns()
 	return nil
 }
 
+// migrateIntervalColumns adds D24 valid_from/valid_to on existing Leaf tables.
+// Fresh CREATE already has them; ALTER is a no-op when the column exists.
+func migrateIntervalColumns() {
+	for _, col := range []string{"valid_from", "valid_to"} {
+		_, _ = conn.Query("ALTER TABLE Leaf ADD " + col + " STRING")
+	}
+}
+
 func closeBrain() {
 	if conn != nil {
 		conn.Close()

internal/brain/http.go

@@ -21,8 +21,8 @@ func Ready() error {
 // HTTP is the in-process API used by bin/brain/serve.go.
 type HTTP struct{}
 
-func (HTTP) Search(ctx context.Context, query string, limit int) ([]byte, error) {
+func (HTTP) Search(ctx context.Context, query string, limit int, asOf string) ([]byte, error) {
-	hits, err := searchHits(query, "", "", limit)
+	hits, err := searchHits(query, "", "", limit, asOf)
 	if err != nil {
 		return nil, err
 	}
@@ -41,7 +41,7 @@ func (HTTP) Search(ctx context.Context, query string, limit int) ([]byte, error)
 	var buf bytes.Buffer
 	enc := json.NewEncoder(&buf)
 	enc.SetEscapeHTML(false)
-	if err := enc.Encode(toJSONOut(hits, query, "", webOut)); err != nil {
+	if err := enc.Encode(toJSONOut(hits, query, "", asOf, webOut)); err != nil {
 		return nil, err
 	}
 	return buf.Bytes(), nil

internal/brain/rank/args.go

@@ -3,12 +3,16 @@ package rank
 import (
 	"fmt"
 	"strconv"
-	"strings"
+
+	"github.com/eSlider/2dph/internal/cli"
+	"github.com/eSlider/2dph/internal/facts"
+	"github.com/integrii/flaggy"
 )
 
-const Usage = `usage: bin/brain/search.go "query" [--root facts|info] [--repo REPO] [-n N] [--hop N] [--json] [--no-web]
+const Usage = `usage: bin/brain/search.go "query" [--root facts|info] [--repo REPO] [-n N] [--hop N] [--as-of YYYY-MM-DD] [--json] [--no-web]
        bin/brain/search.go serve [port]
-       bin/brain/search.go --list-model`
+       bin/brain/search.go --list-model
+       source <(./bin/cli/complete.go bash)`
 
 type Options struct {
 	Query     string
@@ -16,67 +20,71 @@ type Options struct {
 	Repo      string
 	Limit     int
 	Hop       int
+	AsOf      string
 	JSONOut   bool
 	ListModel bool
 	NoWeb     bool
 }
 
+// NewParser is the flaggy schema for search (also used by bin/cli/complete.go).
+func NewParser(opt *Options) *flaggy.Parser {
+	if opt.Limit == 0 {
+		opt.Limit = 20
+	}
+	p := cli.New("brain-search")
+	p.Description = "deduction search: facts → info → web"
+	p.String(&opt.Root, "", "root", "facts or info")
+	p.String(&opt.Repo, "", "repo", "filter by repo")
+	p.Int(&opt.Limit, "n", "n", "max hits")
+	p.Int(&opt.Hop, "", "hop", "walk FROM_FILE depth 1-3")
+	p.String(&opt.AsOf, "", "as-of", "keep facts active on YYYY-MM-DD (D24)")
+	p.Bool(&opt.JSONOut, "", "json", "JSON output")
+	p.Bool(&opt.NoWeb, "", "no-web", "stay local")
+	p.Bool(&opt.ListModel, "", "list-model", "print embedding model")
+	return p
+}
+
 // ParseArgs reads flags. Unknown flags are an error: silently dropping them
 // meant `--hop 1` vanished and its argument `1` was appended to the query.
 func ParseArgs(args []string) (Options, error) {
 	opt := Options{Limit: 20}
-	var queryArgs []string
+	p := NewParser(&opt)
-
+	var q string
-	for i := 0; i < len(args); i++ {
+	p.AddPositionalValue(&q, "query", 1, false, "search query")
-		arg := args[i]
+	if err := cli.Parse(p, args); err != nil {
-		wantsValue := arg == "--root" || arg == "--repo" || arg == "-n" || arg == "--hop"
+		return opt, err
-		if wantsValue && i+1 >= len(args) {
+	}
-			return opt, fmt.Errorf("%s needs a value", arg)
+	opt.Query = cli.Query(q, p.TrailingArguments)
+	if opt.Root != "" && opt.Root != "facts" && opt.Root != "info" {
+		return opt, fmt.Errorf("--root must be facts or info, got %q", opt.Root)
+	}
+	if opt.Limit < 1 {
+		return opt, fmt.Errorf("-n must be a positive integer, got %q", strconv.Itoa(opt.Limit))
+	}
+	if opt.Hop < 0 {
+		return opt, fmt.Errorf("--hop must be a positive integer, got %q", strconv.Itoa(opt.Hop))
+	}
+	if opt.Hop > 3 {
+		return opt, fmt.Errorf("--hop max is 3 (File → Commit → Person)")
+	}
+	if opt.AsOf != "" {
+		day := facts.NormalizeDay(opt.AsOf)
+		if len(day) != 10 || day[4] != '-' || day[7] != '-' {
+			return opt, fmt.Errorf("--as-of must be YYYY-MM-DD, got %q", opt.AsOf)
 		}
-		switch arg {
+		opt.AsOf = day
-		case "--root":
-			i++
-			opt.Root = args[i]
-			if opt.Root != "facts" && opt.Root != "info" {
-				return opt, fmt.Errorf("--root must be facts or info, got %q", opt.Root)
-			}
-		case "--repo":
-			i++
-			opt.Repo = args[i]
-		case "-n":
-			i++
-			n, err := strconv.Atoi(args[i])
-			if err != nil || n < 1 {
-				return opt, fmt.Errorf("-n must be a positive integer, got %q", args[i])
-			}
-			opt.Limit = n
-		case "--hop":
-			i++
-			n, err := strconv.Atoi(args[i])
-			if err != nil || n < 1 {
-				return opt, fmt.Errorf("--hop must be a positive integer, got %q", args[i])
-			}
-			if n > 3 {
-				return opt, fmt.Errorf("--hop max is 3 (File → Commit → Person)")
-			}
-			opt.Hop = n
-		case "--json":
-			opt.JSONOut = true
-		case "--no-web":
-			opt.NoWeb = true
-		case "--list-model":
-			opt.ListModel = true
-		default:
-			if strings.HasPrefix(arg, "-") {
-				return opt, fmt.Errorf("unknown flag %q", arg)
-			}
-			queryArgs = append(queryArgs, arg)
-		}
 	}
-
-	opt.Query = strings.TrimSpace(strings.Join(queryArgs, " "))
 	if opt.Query == "" && !opt.ListModel {
 		return opt, fmt.Errorf("no query given")
 	}
 	return opt, nil
 }
+
+// Parser is the search schema for bin/cli/complete.go.
+func Parser() *flaggy.Parser {
+	opt := Options{Limit: 20}
+	p := NewParser(&opt)
+	var q string
+	p.AddPositionalValue(&q, "query", 1, false, "search query")
+	return p
+}

internal/brain/rank/asof_test.go

@@ -0,0 +1,34 @@
+package rank
+
+import "testing"
+
+func TestFilterAsOfKeepsXDropsY(t *testing.T) {
+	hits := []Hit{
+		{ID: "x", Text: "Andrey works at X", ValidFrom: "2024-03-01", ValidTo: "2025-07-15"},
+		{ID: "y", Text: "Andrey works at Y", ValidFrom: "2025-07-16", ValidTo: ""},
+		{ID: "legacy", Text: "always true claim", ValidFrom: "", ValidTo: ""},
+	}
+	out := FilterAsOf(hits, "2025-01-01")
+	if len(out) != 2 {
+		t.Fatalf("len=%d want 2: %+v", len(out), out)
+	}
+	if out[0].ID != "x" || out[1].ID != "legacy" {
+		t.Fatalf("got %+v", out)
+	}
+	if FilterAsOf(hits, "") == nil || len(FilterAsOf(hits, "")) != 3 {
+		t.Fatal("empty as-of must keep all")
+	}
+}
+
+func TestParseArgsAsOf(t *testing.T) {
+	opt, err := ParseArgs([]string{"who works where", "--as-of", "2025-01-01", "--json"})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if opt.AsOf != "2025-01-01" {
+		t.Fatalf("AsOf=%q", opt.AsOf)
+	}
+	if _, err := ParseArgs([]string{"q", "--as-of", "not-a-date"}); err == nil {
+		t.Fatal("expected bad as-of error")
+	}
+}

internal/brain/rank/escalate.go

@@ -19,20 +19,35 @@ type SecondSourceHit struct {
 
 type WebFn func(query string) SecondSource
 
-// ShouldEscalate is true when the default deduction path has no facts hit.
+// ShouldEscalate is true when the default deduction path has no confirmed
+// facts hit. Hypothesis/partial facts are `(not confirmed)` (D16).
 // `--root facts|info` is a single-root ask: do not mix in the web.
 func ShouldEscalate(hits []Hit, rootFilter string) bool {
 	if rootFilter != "" {
 		return false
 	}
 	for _, h := range hits {
-		if h.Root == "facts" {
+		if ConfirmedFact(h) {
 			return false
 		}
 	}
 	return true
 }
 
+// ConfirmedFact is a facts-root hit that is not hypothesis/partial.
+// Empty confidence is treated as confirmed (legacy leafs).
+func ConfirmedFact(h Hit) bool {
+	if h.Root != "facts" {
+		return false
+	}
+	switch h.Confidence {
+	case "hypothesis", "partial":
+		return false
+	default:
+		return true
+	}
+}
+
 // Deduce returns the second-source block, or nil when web must not run.
 func Deduce(hits []Hit, query, rootFilter string, noWeb bool, web WebFn) *SecondSource {
 	if noWeb || web == nil || !ShouldEscalate(hits, rootFilter) {

internal/brain/rank/escalate_test.go

@@ -14,6 +14,16 @@ func TestShouldEscalateWhenNoFacts(t *testing.T) {
 	}
 }
 
+func TestShouldEscalateWhenHypothesisFacts(t *testing.T) {
+	hyp := Hit{ID: "c", Root: "facts", Confidence: "hypothesis", Source: "a x b vs c x d"}
+	if !ShouldEscalate([]Hit{hyp}, "") {
+		t.Fatal("hypothesis facts are (not confirmed); escalate")
+	}
+	if ConfirmedFact(hyp) {
+		t.Fatal("hypothesis is not confirmed")
+	}
+}
+
 func TestShouldNotEscalateWhenFactsConfirm(t *testing.T) {
 	hits := []Hit{h("f", "facts", "docker ps x compose"), h("i", "info", "docs/a.md")}
 	if ShouldEscalate(hits, "") {

internal/brain/rank/query.go

@@ -3,10 +3,12 @@ package rank
 // BM25 ranks best-first, so the top hits are the *highest* scores; cosine
 // distance ranks best-first ascending. Both mirror kblib.py.
 const FTSStmt = "CALL QUERY_FTS_INDEX('Leaf', 'id', $q) " +
-	"RETURN node.id, node.text, node.root, node.source, score ORDER BY score DESC LIMIT $n"
+	"RETURN node.id, node.text, node.root, node.source, score, node.confidence, " +
+	"node.valid_from, node.valid_to ORDER BY score DESC LIMIT $n"
 
 const VecStmt = "CALL QUERY_VECTOR_INDEX('Leaf', 'Leaf_vec', $q, $n) " +
-	"RETURN node.id, node.text, node.root, node.source, distance ORDER BY distance LIMIT $n"
+	"RETURN node.id, node.text, node.root, node.source, distance, node.confidence, " +
+	"node.valid_from, node.valid_to ORDER BY distance LIMIT $n"
 
 // HopStmt is the Cypher walk from a search hit. Depth 1 = File, 2 = Commit, 3 = Person.
 func HopStmt(depth int) string {

internal/brain/rank/rank.go

@@ -5,6 +5,8 @@ package rank
 import (
 	"sort"
 	"strings"
+
+	"github.com/eSlider/2dph/internal/facts"
 )
 
 type HopNode struct {
@@ -16,23 +18,31 @@ type HopNode struct {
 
 // Hit is one search result, mirroring the python script's dict shape.
 type Hit struct {
-	ID      string    `json:"id"`
+	ID         string    `json:"id"`
-	Text    string    `json:"text"`
+	Text       string    `json:"text"`
-	Root    string    `json:"root"`
+	Root       string    `json:"root"`
-	Source  string    `json:"-"`
+	Confidence string    `json:"confidence,omitempty"`
-	Score   float64   `json:"score"`
+	Source     string    `json:"-"`
-	Snippet string    `json:"snippet,omitempty"`
+	Score      float64   `json:"score"`
-	Hops    []HopNode `json:"hops,omitempty"`
+	Snippet    string    `json:"snippet,omitempty"`
+	ValidFrom  string    `json:"valid_from,omitempty"`
+	ValidTo    string    `json:"valid_to,omitempty"`
+	Hops       []HopNode `json:"hops,omitempty"`
 }
 
 // rrfK dampens the contribution of low ranks; same constant as kblib.py.
 const rrfK = 60
 
-// RankAndFilter fuses the two hit lists, applies --root/--repo, then cuts to
+// RankAndFilter fuses the two hit lists, applies --root/--repo/--as-of, then
-// limit. Cutting first dropped every matching leaf ranked below the cut, so
+// cuts to limit. Cutting first dropped every matching leaf ranked below the
-// `--root facts` came back empty whenever info leafs filled the top N.
+// cut, so `--root facts` came back empty whenever info leafs filled the top N.
-// limit <= 0 keeps everything.
+// limit <= 0 keeps everything. asOf empty skips interval filter (D24).
 func RankAndFilter(fts, vec []Hit, root, repo string, limit int) []Hit {
+	return RankAndFilterAsOf(fts, vec, root, repo, "", limit)
+}
+
+// RankAndFilterAsOf is RankAndFilter with D24 fact-interval filter.
+func RankAndFilterAsOf(fts, vec []Hit, root, repo, asOf string, limit int) []Hit {
 	out := Hybrid(fts, vec, 0)
 	if root != "" {
 		out = FilterRoot(out, root)
@@ -40,12 +50,30 @@ func RankAndFilter(fts, vec []Hit, root, repo string, limit int) []Hit {
 	if repo != "" {
 		out = FilterRepo(out, repo)
 	}
+	if asOf != "" {
+		out = FilterAsOf(out, asOf)
+	}
 	if limit > 0 && len(out) > limit {
 		out = out[:limit]
 	}
 	return out
 }
 
+// FilterAsOf keeps hits whose [valid_from, valid_to] covers asOf (D24).
+// Empty intervals stay (legacy leafs). Empty asOf keeps all.
+func FilterAsOf(hits []Hit, asOf string) []Hit {
+	if asOf == "" {
+		return hits
+	}
+	var out []Hit
+	for _, h := range hits {
+		if facts.ActiveAt(h.ValidFrom, h.ValidTo, asOf) {
+			out = append(out, h)
+		}
+	}
+	return out
+}
+
 // Hybrid merges FTS and vector hits by reciprocal rank fusion.
 // limit <= 0 returns the full fused list.
 func Hybrid(fts, vec []Hit, limit int) []Hit {

internal/brain/rank/rank_test.go

@@ -172,4 +172,7 @@ func TestFTSQueryOrdersByScoreDescending(t *testing.T) {
 	if !strings.Contains(FTSStmt, "ORDER BY score DESC") {
 		t.Fatalf("FTS query must order by score DESC, got:\n%s", FTSStmt)
 	}
+	if !strings.Contains(FTSStmt, "node.confidence") {
+		t.Fatal("FTS must return confidence for D16")
+	}
 }

internal/brain/rank/readargs.go

@@ -0,0 +1,59 @@
+package rank
+
+import (
+	"fmt"
+
+	"github.com/eSlider/2dph/internal/cli"
+	"github.com/integrii/flaggy"
+)
+
+type GetOptions struct {
+	ID      string
+	Body    bool
+	JSONOut bool
+}
+
+func GetParser(opt *GetOptions) *flaggy.Parser {
+	p := cli.New("brain-get")
+	p.Description = "read one leaf"
+	p.Bool(&opt.Body, "", "body", "full text instead of snippet")
+	p.Bool(&opt.JSONOut, "", "json", "JSON output")
+	p.AddPositionalValue(&opt.ID, "id", 1, false, "leaf id")
+	return p
+}
+
+func ParseGet(args []string) (GetOptions, error) {
+	var opt GetOptions
+	if err := cli.Parse(GetParser(&opt), args); err != nil {
+		return opt, err
+	}
+	if opt.ID == "" {
+		return opt, fmt.Errorf("id required")
+	}
+	return opt, nil
+}
+
+type JSONFlag struct {
+	JSONOut bool
+}
+
+func bindJSON(name string, opt *JSONFlag) *flaggy.Parser {
+	p := cli.New(name)
+	p.Bool(&opt.JSONOut, "", "json", "JSON output")
+	return p
+}
+
+func StatsParser() *flaggy.Parser {
+	opt := JSONFlag{}
+	return bindJSON("brain-stats", &opt)
+}
+
+func EvalParser() *flaggy.Parser {
+	opt := JSONFlag{}
+	return bindJSON("brain-eval", &opt)
+}
+
+func ParseJSONFlag(name string, args []string) (JSONFlag, error) {
+	var opt JSONFlag
+	return opt, cli.Parse(bindJSON(name, &opt), args)
+}

internal/brain/read.go

@@ -11,30 +11,15 @@ import (
 	"unicode/utf8"
 
 	"github.com/eSlider/2dph/internal/brain/rank"
+	"github.com/eSlider/2dph/internal/cli"
 )
 
 func MainGet(args []string) int {
-	id, body, jsonOut := "", false, false
+	opt, err := rank.ParseGet(args)
-	for _, a := range args {
+	if err != nil {
-		switch {
+		return cli.Fail(err)
-		case a == "--body":
-			body = true
-		case a == "--json":
-			jsonOut = true
-		case a == "-h" || a == "--help":
-			fmt.Fprintln(os.Stderr, `usage: bin/brain/get.go <id> [--body] [--json]`)
-			return 0
-		case strings.HasPrefix(a, "-"):
-			fmt.Fprintf(os.Stderr, "brain/get: unknown flag %s\n", a)
-			return 2
-		default:
-			id = a
-		}
-	}
-	if id == "" {
-		fmt.Fprintln(os.Stderr, "brain/get: id required")
-		return 2
 	}
+	id, body, jsonOut := opt.ID, opt.Body, opt.JSONOut
 	if err := openBrain(); err != nil {
 		fmt.Fprintf(os.Stderr, "open brain: %v\n", err)
 		return 1
@@ -72,21 +57,11 @@ func MainGet(args []string) int {
 }
 
 func MainStats(args []string) int {
-	jsonOut := false
+	opt, err := rank.ParseJSONFlag("brain-stats", args)
-	for _, a := range args {
+	if err != nil {
-		switch a {
+		return cli.Fail(err)
-		case "--json":
-			jsonOut = true
-		case "-h", "--help":
-			fmt.Fprintln(os.Stderr, `usage: bin/brain/stats.go [--json]`)
-			return 0
-		default:
-			if strings.HasPrefix(a, "-") {
-				fmt.Fprintf(os.Stderr, "brain/stats: unknown flag %s\n", a)
-				return 2
-			}
-		}
 	}
+	jsonOut := opt.JSONOut
 	if err := openBrain(); err != nil {
 		fmt.Fprintf(os.Stderr, "open brain: %v\n", err)
 		return 1
@@ -124,21 +99,11 @@ func MainStats(args []string) int {
 }
 
 func MainEval(args []string) int {
-	jsonOut := false
+	opt, err := rank.ParseJSONFlag("brain-eval", args)
-	for _, a := range args {
+	if err != nil {
-		switch a {
+		return cli.Fail(err)
-		case "--json":
-			jsonOut = true
-		case "-h", "--help":
-			fmt.Fprintln(os.Stderr, `usage: bin/brain/eval.go [--json]`)
-			return 0
-		default:
-			if strings.HasPrefix(a, "-") {
-				fmt.Fprintf(os.Stderr, "brain/eval: unknown flag %s\n", a)
-				return 2
-			}
-		}
 	}
+	jsonOut := opt.JSONOut
 	if err := openBrain(); err != nil {
 		fmt.Fprintf(os.Stderr, "open brain: %v\n", err)
 		return 1

internal/brain/search.go

@@ -20,6 +20,7 @@ import (
 
 	lbug "github.com/LadybugDB/go-ladybug"
 	"github.com/eSlider/2dph/internal/brain/rank"
+	"github.com/eSlider/2dph/internal/cli"
 )
 
 const defaultPort = 17830
@@ -29,6 +30,9 @@ const healthPath = "/health"
 func runSearch(args []string) int {
 	opt, err := rank.ParseArgs(args)
 	if err != nil {
+		if errors.Is(err, cli.ErrHelp) {
+			return 0
+		}
 		fmt.Fprintf(os.Stderr, "brain/search: %v\n%s\n", err, rank.Usage)
 		return 2
 	}
@@ -51,7 +55,7 @@ func runSearch(args []string) int {
 	}
 	defer closeBrain()
 
-	hits, err := searchHits(query, root, repo, limit)
+	hits, err := searchHits(query, root, repo, limit, opt.AsOf)
 	if err != nil {
 		fmt.Fprintf(os.Stderr, "search: %v\n", err)
 		return 1
@@ -81,6 +85,7 @@ func runSearch(args []string) int {
 	out := Dict{
 		{"query", query},
 		{"root_filter", root},
+		{"as_of", opt.AsOf},
 		{"count", len(results)},
 		{"results", resultsToDicts(results)},
 	}
@@ -92,13 +97,13 @@ func runSearch(args []string) int {
 		enc := json.NewEncoder(os.Stdout)
 		enc.SetIndent("", "  ")
 		enc.SetEscapeHTML(false)
-		return b2i(enc.Encode(toJSONOut(results, query, root, webOut)))
+		return b2i(enc.Encode(toJSONOut(results, query, root, opt.AsOf, webOut)))
 	}
 	fmt.Print(toYAML(out, 0))
 	return 0
 }
 
-func searchHits(query, root, repo string, limit int) ([]Hit, error) {
+func searchHits(query, root, repo string, limit int, asOf string) ([]Hit, error) {
 	emb, err := embedQuery(query)
 	if err != nil {
 		return nil, fmt.Errorf("embed: %w", err)
@@ -111,7 +116,7 @@ func searchHits(query, root, repo string, limit int) ([]Hit, error) {
 	if vec, err = queryVector(emb, limit*3); err != nil {
 		fmt.Fprintf(os.Stderr, "vec: %v\n", err)
 	}
-	return rank.RankAndFilter(fts, vec, root, repo, limit), nil
+	return rank.RankAndFilterAsOf(fts, vec, root, repo, asOf, limit), nil
 }
 
 func attachHops(hits []Hit, n int) error {
@@ -212,44 +217,75 @@ func rowsToHits(res *lbug.QueryResult) ([]Hit, error) {
 		root := fmt.Sprint(vals[2])
 		source := fmt.Sprint(vals[3])
 		score := float64(vals[4].(float64))
-		hits = append(hits, Hit{ID: id, Text: text, Root: root, Source: source, Score: score})
+		conf := ""
+		if len(vals) >= 6 {
+			conf = fmt.Sprint(vals[5])
+		}
+		vf, vt := "", ""
+		if len(vals) >= 8 {
+			vf = nullStr(vals[6])
+			vt = nullStr(vals[7])
+		}
+		hits = append(hits, Hit{
+			ID: id, Text: text, Root: root, Source: source, Score: score,
+			Confidence: conf, ValidFrom: vf, ValidTo: vt,
+		})
 	}
 	return hits, nil
 }
 
+func nullStr(v any) string {
+	if v == nil {
+		return ""
+	}
+	s := fmt.Sprint(v)
+	if s == "<nil>" {
+		return ""
+	}
+	return s
+}
+
 // JSON output types
 type jsonOut struct {
 	Query      string             `json:"query"`
 	RootFilter string             `json:"root_filter"`
+	AsOf       string             `json:"as_of,omitempty"`
 	Count      int                `json:"count"`
 	Results    []jsonHit          `json:"results"`
 	Web        *rank.SecondSource `json:"web,omitempty"`
 }
 
 type jsonHit struct {
-	ID      string         `json:"id"`
+	ID         string         `json:"id"`
-	Text    string         `json:"text"`
+	Text       string         `json:"text"`
-	Root    string         `json:"root"`
+	Root       string         `json:"root"`
-	Score   float64        `json:"score"`
+	Confidence string         `json:"confidence,omitempty"`
-	Snippet string         `json:"snippet,omitempty"`
+	Score      float64        `json:"score"`
-	Hops    []rank.HopNode `json:"hops,omitempty"`
+	Snippet    string         `json:"snippet,omitempty"`
+	ValidFrom  string         `json:"valid_from,omitempty"`
+	ValidTo    string         `json:"valid_to,omitempty"`
+	Hops       []rank.HopNode `json:"hops,omitempty"`
 }
 
-func toJSONOut(hits []Hit, query, rootFilter string, web *rank.SecondSource) *jsonOut {
+func toJSONOut(hits []Hit, query, rootFilter, asOf string, web *rank.SecondSource) *jsonOut {
 	out := make([]jsonHit, len(hits))
 	for i, h := range hits {
 		out[i] = jsonHit{
-			ID:      h.ID,
+			ID:         h.ID,
-			Text:    h.Text,
+			Text:       h.Text,
-			Root:    h.Root,
+			Root:       h.Root,
-			Score:   h.Score,
+			Confidence: h.Confidence,
-			Snippet: h.Snippet,
+			Score:      h.Score,
-			Hops:    h.Hops,
+			Snippet:    h.Snippet,
+			ValidFrom:  h.ValidFrom,
+			ValidTo:    h.ValidTo,
+			Hops:       h.Hops,
 		}
 	}
 	return &jsonOut{
 		Query:      query,
 		RootFilter: rootFilter,
+		AsOf:       asOf,
 		Count:      len(hits),
 		Results:    out,
 		Web:        web,
@@ -265,6 +301,15 @@ func resultsToDicts(hits []Hit) []any {
 			{"root", h.Root},
 			{"score", h.Score},
 		}
+		if h.Confidence != "" {
+			d = append(d, KV{"confidence", h.Confidence})
+		}
+		if h.ValidFrom != "" {
+			d = append(d, KV{"valid_from", h.ValidFrom})
+		}
+		if h.ValidTo != "" {
+			d = append(d, KV{"valid_to", h.ValidTo})
+		}
 		if h.Snippet != "" {
 			d = append(d, KV{"snippet", h.Snippet})
 		}

internal/chats/apply.go

@@ -3,21 +3,22 @@ package chats
 import (
 	"bytes"
 	"encoding/json"
-	"flag"
 	"fmt"
 	"os"
 	"os/exec"
 	"path/filepath"
 	"strings"
+
+	cliparse "github.com/eSlider/2dph/internal/cli"
 )
 
 type ooContact struct {
-	ID         int    `json:"id"`
+	ID          int    `json:"id"`
 	DisplayName string `json:"displayName"`
-	FirstName  string `json:"firstName"`
+	FirstName   string `json:"firstName"`
-	LastName   string `json:"lastName"`
+	LastName    string `json:"lastName"`
-	About      string `json:"about"`
+	About       string `json:"about"`
-	CommonData []struct {
+	CommonData  []struct {
 		InfoType int    `json:"infoType"`
 		Data     string `json:"data"`
 		Category string `json:"categoryName"`
@@ -25,16 +26,9 @@ type ooContact struct {
 }
 
 func RunApply(args []string) int {
-	fs := flag.NewFlagSet("chats apply", flag.ContinueOnError)
+	dryRun, err := parseApplyFlags(args)
-	dryRun := fs.Bool("dry-run", false, "show what would be done without writing")
+	if err != nil {
-	help := fs.Bool("help", false, "")
+		return cliparse.Fail(err)
-	fs.SetOutput(os.Stderr)
-	if err := fs.Parse(args); err != nil {
-		return 2
-	}
-	if *help {
-		fmt.Fprintln(os.Stderr, "usage: chats apply [--dry-run]")
-		return 0
 	}
 
 	ooCLI := findOO()
@@ -60,10 +54,10 @@ func RunApply(args []string) int {
 	emailFacts = dedupeFacts(emailFacts)
 
 	type resolvedFact struct {
-		Fact       ExtractedFact
+		Fact   ExtractedFact
-		OoID       int
+		OoID   int
-		OoName     string
+		OoName string
-		Action     string // "info-add" or "persons-create"
+		Action string // "info-add" or "persons-create"
 	}
 
 	var resolved []resolvedFact
@@ -128,7 +122,7 @@ func RunApply(args []string) int {
 
 	fmt.Printf("\nchats apply: %d actions to apply\n", len(resolved))
 
-	if *dryRun {
+	if dryRun {
 		for _, r := range resolved {
 			switch r.Action {
 			case "info-add":

internal/chats/cli.go

@@ -0,0 +1,74 @@
+package chats
+
+import (
+	cliparse "github.com/eSlider/2dph/internal/cli"
+	"github.com/integrii/flaggy"
+)
+
+type syncTelegramFlags struct {
+	Limit int
+	Phone string
+}
+
+type syncLinkedInFlags struct {
+	Limit   int
+	Refresh bool
+}
+
+func SyncParser() *flaggy.Parser {
+	p := cliparse.New("chats-sync")
+	p.Description = "download chats to var/chats"
+	tg := flaggy.NewSubcommand("telegram")
+	li := flaggy.NewSubcommand("linkedin")
+	var limit int
+	var phone string
+	var refresh bool
+	tg.Int(&limit, "", "limit", "max messages per chat")
+	tg.String(&phone, "", "phone", "phone (default TELEGRAM_PHONE)")
+	li.Int(&limit, "", "limit", "max messages per conversation")
+	li.Bool(&refresh, "", "refresh", "refresh webtop session")
+	p.AttachSubcommand(tg, 1)
+	p.AttachSubcommand(li, 1)
+	return p
+}
+
+func ImportParser() *flaggy.Parser {
+	return cliparse.New("chats-import")
+}
+
+func FactsParser() *flaggy.Parser {
+	return cliparse.New("chats-facts")
+}
+
+func ApplyParser() *flaggy.Parser {
+	p := cliparse.New("chats-apply")
+	dry := false
+	p.Bool(&dry, "", "dry-run", "show without writing")
+	return p
+}
+
+func parseTelegramFlags(args []string) (syncTelegramFlags, error) {
+	var f syncTelegramFlags
+	p := cliparse.New("chats-sync-telegram")
+	p.Int(&f.Limit, "", "limit", "max messages per chat")
+	p.String(&f.Phone, "", "phone", "phone (default TELEGRAM_PHONE)")
+	return f, cliparse.Parse(p, args)
+}
+
+func parseLinkedInFlags(args []string) (syncLinkedInFlags, error) {
+	var f syncLinkedInFlags
+	p := cliparse.New("chats-sync-linkedin")
+	p.Int(&f.Limit, "", "limit", "max messages per conversation")
+	p.Bool(&f.Refresh, "", "refresh", "refresh webtop session")
+	return f, cliparse.Parse(p, args)
+}
+
+func parseApplyFlags(args []string) (dryRun bool, err error) {
+	p := cliparse.New("chats-apply")
+	p.Bool(&dryRun, "", "dry-run", "show without writing")
+	return dryRun, cliparse.Parse(p, args)
+}
+
+func parseNoFlags(name string, args []string) error {
+	return cliparse.Parse(cliparse.New(name), args)
+}

internal/chats/facts.go

@@ -3,12 +3,13 @@ package chats
 import (
 	"bufio"
 	"encoding/json"
-	"flag"
 	"fmt"
 	"os"
 	"path/filepath"
 	"regexp"
 	"strings"
+
+	cliparse "github.com/eSlider/2dph/internal/cli"
 )
 
 var (
@@ -76,15 +77,8 @@ type ExtractedFact struct {
 }
 
 func RunFacts(args []string) int {
-	fs := flag.NewFlagSet("chats facts", flag.ContinueOnError)
+	if err := parseNoFlags("chats-facts", args); err != nil {
-	help := fs.Bool("help", false, "")
+		return cliparse.Fail(err)
-	fs.SetOutput(os.Stderr)
-	if err := fs.Parse(args); err != nil {
-		return 2
-	}
-	if *help {
-		fmt.Fprintln(os.Stderr, "usage: chats facts")
-		return 0
 	}
 
 	root := Dir()

internal/chats/import.go

@@ -4,25 +4,19 @@ import (
 	"bufio"
 	"bytes"
 	"encoding/json"
-	"flag"
 	"fmt"
 	"html"
 	"os"
 	"path/filepath"
 	"sort"
 	"strings"
+
+	cliparse "github.com/eSlider/2dph/internal/cli"
 )
 
 func RunImport(args []string) int {
-	fs := flag.NewFlagSet("chats import", flag.ContinueOnError)
+	if err := parseNoFlags("chats-import", args); err != nil {
-	help := fs.Bool("help", false, "")
+		return cliparse.Fail(err)
-	fs.SetOutput(os.Stderr)
-	if err := fs.Parse(args); err != nil {
-		return 2
-	}
-	if *help {
-		fmt.Fprintln(os.Stderr, "usage: chats import")
-		return 0
 	}
 
 	root := Dir()

internal/chats/sync_linkedin.go

@@ -2,12 +2,13 @@ package chats
 
 import (
 	"context"
-	"flag"
 	"fmt"
 	"os"
 	"os/exec"
 	"path/filepath"
 	"time"
+
+	cliparse "github.com/eSlider/2dph/internal/cli"
 )
 
 func checkLinkedInSession(userDataDir string) (bool, error) {
@@ -29,18 +30,12 @@ func checkLinkedInSession(userDataDir string) (bool, error) {
 }
 
 func RunSyncLinkedIn(args []string) int {
-	fs := flag.NewFlagSet("chats sync linkedin", flag.ContinueOnError)
+	f, err := parseLinkedInFlags(args)
-	limit := fs.Int("limit", 0, "max messages per conversation (0 = all)")
+	if err != nil {
-	refresh := fs.Bool("refresh", false, "refresh session from live webtop browser before sync")
+		return cliparse.Fail(err)
-	help := fs.Bool("help", false, "")
-	fs.SetOutput(os.Stderr)
-	if err := fs.Parse(args); err != nil {
-		return 2
-	}
-	if *help {
-		fmt.Fprintln(os.Stderr, "usage: chats sync linkedin [--limit N] [--refresh]")
-		return 0
 	}
+	limit := f.Limit
+	refresh := f.Refresh
 
 	userDataDir := envVar("LINKEDIN_USER_DATA_DIR", "")
 	if userDataDir == "" {
@@ -48,7 +43,7 @@ func RunSyncLinkedIn(args []string) int {
 		userDataDir = home + "/.linkedin-mcp/profile"
 	}
 
-	if *refresh {
+	if refresh {
 		if code := refreshLinkedInSession(userDataDir); code != 0 {
 			return code
 		}
@@ -72,7 +67,7 @@ func RunSyncLinkedIn(args []string) int {
 	defer cancel()
 
 	start := time.Now()
-	if err := src.Sync(ctx, Dir(), *limit); err != nil {
+	if err := src.Sync(ctx, Dir(), limit); err != nil {
 		fmt.Fprintf(os.Stderr, "chats sync linkedin: %v\n", err)
 		return 1
 	}

internal/chats/sync_telegram.go

@@ -2,33 +2,28 @@ package chats
 
 import (
 	"context"
-	"flag"
 	"fmt"
 	"os"
 	"path/filepath"
 	"strconv"
 	"strings"
 	"time"
+
+	cliparse "github.com/eSlider/2dph/internal/cli"
 )
 
 func RunSyncTelegram(args []string) int {
-	fs := flag.NewFlagSet("chats sync telegram", flag.ContinueOnError)
+	f, err := parseTelegramFlags(args)
-	limit := fs.Int("limit", 0, "max messages per chat (0 = all)")
+	if err != nil {
-	phone := fs.String("phone", "", "phone number (default env TELEGRAM_PHONE)")
+		return cliparse.Fail(err)
-	help := fs.Bool("help", false, "")
-	fs.SetOutput(os.Stderr)
-	if err := fs.Parse(args); err != nil {
-		return 2
-	}
-	if *help {
-		fmt.Fprintln(os.Stderr, "usage: chats sync telegram [--limit N] [--phone PHONE]")
-		return 0
 	}
+	limit := f.Limit
+	phone := f.Phone
 
 	apiIDStr := envVar("TELEGRAM_API_ID", "")
 	apiHash := envVar("TELEGRAM_API_HASH", "")
 	sessionStr := envVar("TELEGRAM_SESSION_STRING", "")
-	phoneNum := *phone
+	phoneNum := phone
 	if phoneNum == "" {
 		phoneNum = envVar("TELEGRAM_PHONE", "")
 	}
@@ -78,7 +73,7 @@ func RunSyncTelegram(args []string) int {
 	defer cancel()
 
 	start := time.Now()
-	if err := src.Sync(ctx, Dir(), *limit); err != nil {
+	if err := src.Sync(ctx, Dir(), limit); err != nil {
 		fmt.Fprintf(os.Stderr, "chats sync telegram: %v\n", err)
 		return 1
 	}

internal/cli/cli.go

@@ -0,0 +1,190 @@
+// Package cli is the shared flaggy wrapper (D23).
+//
+// flaggy: zero deps, flags at any position, shell completion scripts.
+// Individual tools keep ShowCompletion off so a query like "completion" is
+// not stolen; dump scripts with bin/cli/complete.go.
+package cli
+
+import (
+	"errors"
+	"fmt"
+	"os"
+	"strings"
+	"sync"
+
+	"github.com/integrii/flaggy"
+)
+
+// ErrHelp means -h/--help was requested (exit 0).
+var ErrHelp = errors.New("help")
+
+var parseMu sync.Mutex
+
+// New returns a per-call parser. Never reuse: flaggy parses once.
+func New(name string) *flaggy.Parser {
+	p := flaggy.NewParser(name)
+	p.ShowVersionWithVersionFlag = false
+	p.ShowCompletion = false
+	// Extra positionals become TrailingArguments (search "two words --json").
+	// Unknown dash tokens are rejected in Parse after flaggy returns.
+	p.ShowHelpOnUnexpected = false
+	p.ShowHelpWithHFlag = true
+	return p
+}
+
+// Parse runs p.ParseArgs and turns flaggy's os.Exit into an error.
+// Not safe to call in parallel (flaggy.PanicInsteadOfExit is process-global).
+func Parse(p *flaggy.Parser, args []string) error {
+	parseMu.Lock()
+	defer parseMu.Unlock()
+	prev := flaggy.PanicInsteadOfExit
+	flaggy.PanicInsteadOfExit = true
+	defer func() { flaggy.PanicInsteadOfExit = prev }()
+
+	var exitMsg string
+	err := func() error {
+		defer func() {
+			if r := recover(); r != nil {
+				exitMsg = fmt.Sprint(r)
+			}
+		}()
+		return p.ParseArgs(args)
+	}()
+	if err != nil {
+		return err
+	}
+	if exitMsg != "" {
+		if strings.Contains(exitMsg, "code: 0") {
+			return ErrHelp
+		}
+		return errors.New(exitMsg)
+	}
+	if u := unknownFlags(p, args); len(u) > 0 {
+		return fmt.Errorf("unknown flag %q", u[0])
+	}
+	return nil
+}
+
+func unknownFlags(p *flaggy.Parser, args []string) []string {
+	flags := collectFlags(&p.Subcommand)
+	var out []string
+	skipNext := false
+	for _, a := range args {
+		if skipNext {
+			skipNext = false
+			continue
+		}
+		if a == "--" {
+			break
+		}
+		name, inline := flagName(a)
+		if name == "" {
+			continue
+		}
+		if name == "h" || name == "help" {
+			continue
+		}
+		f := findFlag(flags, name)
+		if f == nil {
+			out = append(out, a)
+			continue
+		}
+		if !inline && !isBoolFlag(f) {
+			skipNext = true
+		}
+	}
+	return out
+}
+
+func flagName(a string) (name string, inline bool) {
+	if a == "-" || !strings.HasPrefix(a, "-") {
+		return "", false
+	}
+	rest := strings.TrimLeft(a, "-")
+	name, _, inline = strings.Cut(rest, "=")
+	return name, inline
+}
+
+func collectFlags(sc *flaggy.Subcommand) []*flaggy.Flag {
+	out := append([]*flaggy.Flag{}, sc.Flags...)
+	for _, sub := range sc.Subcommands {
+		out = append(out, collectFlags(sub)...)
+	}
+	return out
+}
+
+func findFlag(flags []*flaggy.Flag, name string) *flaggy.Flag {
+	for _, f := range flags {
+		if f.HasName(name) {
+			return f
+		}
+	}
+	return nil
+}
+
+func isBoolFlag(f *flaggy.Flag) bool {
+	_, ok := f.AssignmentVar.(*bool)
+	return ok
+}
+
+// Query joins the first positional with leftover trailing words.
+func Query(first string, trailing []string) string {
+	parts := make([]string, 0, 1+len(trailing))
+	if s := strings.TrimSpace(first); s != "" {
+		parts = append(parts, s)
+	}
+	for _, t := range trailing {
+		if s := strings.TrimSpace(t); s != "" {
+			parts = append(parts, s)
+		}
+	}
+	return strings.Join(parts, " ")
+}
+
+// Code maps parse errors to process exit codes (0 help, 2 usage).
+func Code(err error) int {
+	if err == nil || errors.Is(err, ErrHelp) {
+		return 0
+	}
+	return 2
+}
+
+// Fail prints err unless it is help or a flaggy exit that already wrote stderr.
+func Fail(err error) int {
+	if err == nil || errors.Is(err, ErrHelp) {
+		return 0
+	}
+	if strings.HasPrefix(err.Error(), "Panic instead of exit") {
+		return 2
+	}
+	fmt.Fprintln(os.Stderr, err)
+	return 2
+}
+
+// Tool is one shebang CLI for completion dump.
+type Tool struct {
+	Path string
+	Name string
+	New  func() *flaggy.Parser
+}
+
+// BashScript concatenates flaggy bash complete scripts and binds each
+// function to the shebang path (./bin/subject/method.go).
+func BashScript(tools []Tool) string {
+	var b strings.Builder
+	b.WriteString("# 2dph flaggy completions (D23). source <(./bin/cli/complete.go bash)\n")
+	for _, t := range tools {
+		p := t.New()
+		p.Name = t.Name
+		script := flaggy.GenerateBashCompletion(p)
+		b.WriteString(script)
+		fn := "_" + strings.ReplaceAll(t.Name, "-", "_") + "_complete"
+		if t.Path != "" && t.Path != t.Name {
+			fmt.Fprintf(&b, "complete -F %s %s\n", fn, t.Path)
+			if !strings.HasPrefix(t.Path, "./") {
+				fmt.Fprintf(&b, "complete -F %s ./%s\n", fn, t.Path)
+			}
+		}
+	}
+	return b.String()
+}

internal/cli/cli_test.go

@@ -0,0 +1,77 @@
+package cli
+
+import (
+	"errors"
+	"strings"
+	"testing"
+
+	"github.com/integrii/flaggy"
+)
+
+func TestParseBoolAndIntAnyPosition(t *testing.T) {
+	p := New("t")
+	jsonOut := false
+	n := 20
+	q := ""
+	p.Bool(&jsonOut, "", "json", "JSON")
+	p.Int(&n, "n", "n", "limit")
+	p.AddPositionalValue(&q, "query", 1, false, "q")
+	if err := Parse(p, []string{"two", "words", "--json", "-n", "5"}); err != nil {
+		t.Fatal(err)
+	}
+	got := Query(q, p.TrailingArguments)
+	if got != "two words" || !jsonOut || n != 5 {
+		t.Fatalf("q=%q json=%v n=%d", got, jsonOut, n)
+	}
+}
+
+func TestParseUnknownFlagIsError(t *testing.T) {
+	p := New("t")
+	jsonOut := false
+	p.Bool(&jsonOut, "", "json", "JSON")
+	if err := Parse(p, []string{"--nope"}); err == nil {
+		t.Fatal("unknown flag accepted")
+	}
+}
+
+func TestParseHelpIsErrHelp(t *testing.T) {
+	p := New("t")
+	jsonOut := false
+	p.Bool(&jsonOut, "", "json", "JSON")
+	err := Parse(p, []string{"--help"})
+	if !errors.Is(err, ErrHelp) {
+		t.Fatalf("got %v", err)
+	}
+}
+
+func TestParseMissingFlagValueIsError(t *testing.T) {
+	p := New("t")
+	n := 0
+	p.Int(&n, "", "hop", "hop")
+	if err := Parse(p, []string{"--hop"}); err == nil {
+		t.Fatal("expected missing value error")
+	}
+}
+
+func TestBashScriptNamesShebangPath(t *testing.T) {
+	out := BashScript([]Tool{{
+		Path: "bin/brain/search.go",
+		Name: "brain-search",
+		New:  newSearchLike,
+	}})
+	if !strings.Contains(out, "--json") || !strings.Contains(out, "--hop") {
+		t.Fatalf("flags missing:\n%s", out)
+	}
+	if !strings.Contains(out, "complete -F") || !strings.Contains(out, "bin/brain/search.go") {
+		t.Fatalf("shebang complete missing:\n%s", out)
+	}
+}
+
+func newSearchLike() *flaggy.Parser {
+	p := New("brain-search")
+	jsonOut := false
+	hop := 0
+	p.Bool(&jsonOut, "", "json", "JSON")
+	p.Int(&hop, "", "hop", "graph hop")
+	return p
+}

internal/cli/qa.go

@@ -0,0 +1,24 @@
+package cli
+
+import "github.com/integrii/flaggy"
+
+type QAStats struct {
+	JSONL string
+}
+
+func QAParser() *flaggy.Parser {
+	c := QAStats{}
+	return BindQA(&c)
+}
+
+func BindQA(c *QAStats) *flaggy.Parser {
+	p := New("qa-stats")
+	p.Description = "DuckDB quantiles / JSONL count"
+	p.String(&c.JSONL, "", "jsonl", "JSONL file (else stdin JSON [float,…])")
+	return p
+}
+
+func ParseQAStats(args []string) (QAStats, error) {
+	var c QAStats
+	return c, Parse(BindQA(&c), args)
+}

internal/facts/contradict.go

@@ -0,0 +1,120 @@
+// Package facts is cgo-free evidence rules (D16 contradictions).
+package facts
+
+import "strconv"
+
+const (
+	ConfConfirmed  = "confirmed"
+	ConfHypothesis = "hypothesis"
+
+	RuleUnresolved        = "unresolved"
+	RuleTemporalFreshness = "temporal_freshness"
+	RuleAuthorityPairing  = "authority_pairing"
+	RuleTwoSource         = "two_source"
+	RuleSingleSource      = "single_source"
+
+	KindRuntime   = "runtime"
+	KindConfig    = "config"
+	KindNarrative = "narrative"
+)
+
+// Source is one independent pointer on a yes or no side.
+type Source struct {
+	ID    string `json:"id"`
+	Kind  string `json:"kind"`
+	When  string `json:"when,omitempty"`
+	Stale bool   `json:"stale,omitempty"`
+}
+
+// Claim is one assertion with yes/no evidence lists.
+type Claim struct {
+	Text string   `json:"text"`
+	Yes  []Source `json:"yes"`
+	No   []Source `json:"no"`
+}
+
+// Result is audit output. Confirmed=false means `(not confirmed)`.
+type Result struct {
+	Text       string `json:"text"`
+	Confidence string `json:"confidence"`
+	Confirmed  bool   `json:"confirmed"`
+	Rule       string `json:"rule"`
+	Winner     string `json:"winner,omitempty"`
+	YesN       int    `json:"yes"`
+	NoN        int    `json:"no"`
+}
+
+func independent(ss []Source) int {
+	seen := map[string]struct{}{}
+	for i, s := range ss {
+		id := s.ID
+		if id == "" {
+			id = s.Kind + "#" + strconv.Itoa(i)
+		}
+		seen[id] = struct{}{}
+	}
+	return len(seen)
+}
+
+func freshN(ss []Source) int {
+	n := 0
+	for _, s := range ss {
+		if !s.Stale {
+			n++
+		}
+	}
+	return n
+}
+
+func strongN(ss []Source) int {
+	n := 0
+	for _, s := range ss {
+		if s.Kind == KindRuntime || s.Kind == KindConfig {
+			n++
+		}
+	}
+	return n
+}
+
+func out(c Claim, conf, rule, winner string) Result {
+	return Result{
+		Text:       c.Text,
+		Confidence: conf,
+		Confirmed:  conf == ConfConfirmed,
+		Rule:       rule,
+		Winner:     winner,
+		YesN:       independent(c.Yes),
+		NoN:        independent(c.No),
+	}
+}
+
+// Adjudicate applies D16: ≥2 yes vs ≥2 no stays hypothesis until a rule fires.
+// Order: temporal_freshness, then authority_pairing (A/B beats narrative C).
+func Adjudicate(c Claim) Result {
+	yesN := independent(c.Yes)
+	noN := independent(c.No)
+	if yesN < 2 || noN < 2 {
+		if yesN >= 2 {
+			return out(c, ConfConfirmed, RuleTwoSource, "yes")
+		}
+		if noN >= 2 {
+			return out(c, ConfConfirmed, RuleTwoSource, "no")
+		}
+		return out(c, ConfHypothesis, RuleSingleSource, "")
+	}
+	yf, nf := freshN(c.Yes), freshN(c.No)
+	if yf >= 2 && nf < 2 {
+		return out(c, ConfConfirmed, RuleTemporalFreshness, "yes")
+	}
+	if nf >= 2 && yf < 2 {
+		return out(c, ConfConfirmed, RuleTemporalFreshness, "no")
+	}
+	ys, ns := strongN(c.Yes), strongN(c.No)
+	if ys >= 2 && ns < 2 {
+		return out(c, ConfConfirmed, RuleAuthorityPairing, "yes")
+	}
+	if ns >= 2 && ys < 2 {
+		return out(c, ConfConfirmed, RuleAuthorityPairing, "no")
+	}
+	return out(c, ConfHypothesis, RuleUnresolved, "")
+}

internal/facts/contradict_test.go

@@ -0,0 +1,86 @@
+package facts
+
+import "testing"
+
+func src(id, kind string, stale bool) Source {
+	return Source{ID: id, Kind: kind, Stale: stale}
+}
+
+func TestTwoVsTwoStaysHypothesis(t *testing.T) {
+	c := Claim{
+		Text: "svc listens on 443",
+		Yes: []Source{
+			src("docker-ps", KindRuntime, false),
+			src("compose", KindConfig, false),
+		},
+		No: []Source{
+			src("docker-ps-old", KindRuntime, false),
+			src("compose-old", KindConfig, false),
+		},
+	}
+	r := Adjudicate(c)
+	if r.Confirmed || r.Confidence != ConfHypothesis || r.Rule != RuleUnresolved {
+		t.Fatalf("2v2 must stay (not confirmed): %+v", r)
+	}
+	if r.Winner != "" {
+		t.Fatalf("unresolved must not name a winner: %+v", r)
+	}
+}
+
+func TestTemporalFreshnessResolvesStaleSide(t *testing.T) {
+	c := Claim{
+		Text: "svc listens on 443",
+		Yes: []Source{
+			src("docker-ps", KindRuntime, false),
+			src("compose", KindConfig, false),
+		},
+		No: []Source{
+			src("old-readme", KindNarrative, true),
+			src("old-wiki", KindNarrative, true),
+		},
+	}
+	r := Adjudicate(c)
+	if !r.Confirmed || r.Rule != RuleTemporalFreshness || r.Winner != "yes" {
+		t.Fatalf("fresh yes vs stale no: %+v", r)
+	}
+}
+
+func TestAuthorityPairingBeatsNarrative(t *testing.T) {
+	c := Claim{
+		Text: "svc listens on 443",
+		Yes: []Source{
+			src("docker-ps", KindRuntime, false),
+			src("compose", KindConfig, false),
+		},
+		No: []Source{
+			src("readme", KindNarrative, false),
+			src("wiki", KindNarrative, false),
+		},
+	}
+	r := Adjudicate(c)
+	if !r.Confirmed || r.Rule != RuleAuthorityPairing || r.Winner != "yes" {
+		t.Fatalf("A×B vs C×C: %+v", r)
+	}
+}
+
+func TestTwoSourceYesIsConfirmed(t *testing.T) {
+	c := Claim{
+		Text: "arc-1 runs Matrix",
+		Yes: []Source{
+			src("compose", KindConfig, false),
+			src("docker-ps", KindRuntime, false),
+		},
+	}
+	r := Adjudicate(c)
+	if !r.Confirmed || r.Rule != RuleTwoSource || r.Winner != "yes" {
+		t.Fatalf("%+v", r)
+	}
+}
+
+func TestSingleSourceIsHypothesis(t *testing.T) {
+	c := Claim{Text: "maybe", Yes: []Source{src("readme", KindNarrative, false)}}
+	r := Adjudicate(c)
+	if r.Confirmed || r.Rule != RuleSingleSource {
+		t.Fatalf("%+v", r)
+	}
+}

internal/facts/interval.go

@@ -0,0 +1,44 @@
+package facts
+
+// Interval of truth for a fact leaf (D24 / OQ5). Not D16 source staleness.
+//
+// Empty valid_from and valid_to means "always" (legacy leafs). Empty asOf
+// means "do not filter". Dates compare as YYYY-MM-DD (lexicographic).
+
+// NormalizeDay keeps the calendar day from ISO-8601 or bare dates.
+func NormalizeDay(s string) string {
+	s = trimSpace(s)
+	if len(s) >= 10 && s[4] == '-' && s[7] == '-' {
+		return s[:10]
+	}
+	return s
+}
+
+func trimSpace(s string) string {
+	i, j := 0, len(s)
+	for i < j && (s[i] == ' ' || s[i] == '\t' || s[i] == '\n' || s[i] == '\r') {
+		i++
+	}
+	for j > i && (s[j-1] == ' ' || s[j-1] == '\t' || s[j-1] == '\n' || s[j-1] == '\r') {
+		j--
+	}
+	return s[i:j]
+}
+
+// ActiveAt reports whether a fact with [validFrom, validTo] holds at asOf.
+// validTo empty = open-ended. Both ends inclusive.
+func ActiveAt(validFrom, validTo, asOf string) bool {
+	asOf = NormalizeDay(asOf)
+	if asOf == "" {
+		return true
+	}
+	from := NormalizeDay(validFrom)
+	to := NormalizeDay(validTo)
+	if from != "" && asOf < from {
+		return false
+	}
+	if to != "" && asOf > to {
+		return false
+	}
+	return true
+}

internal/facts/interval_test.go

@@ -0,0 +1,73 @@
+package facts
+
+import "testing"
+
+func TestActiveAtOpenEnded(t *testing.T) {
+	// works at Y from 2025-07-16, no end
+	if !ActiveAt("2025-07-16", "", "2025-07-16") {
+		t.Fatal("inclusive valid_from")
+	}
+	if !ActiveAt("2025-07-16", "", "2026-01-01") {
+		t.Fatal("open-ended valid_to")
+	}
+	if ActiveAt("2025-07-16", "", "2025-07-15") {
+		t.Fatal("before valid_from must be inactive")
+	}
+}
+
+func TestActiveAtClosedInterval(t *testing.T) {
+	// works at X 2024-03-01 .. 2025-07-15
+	if !ActiveAt("2024-03-01", "2025-07-15", "2025-01-01") {
+		t.Fatal("mid interval")
+	}
+	if !ActiveAt("2024-03-01", "2025-07-15", "2024-03-01") {
+		t.Fatal("inclusive start")
+	}
+	if !ActiveAt("2024-03-01", "2025-07-15", "2025-07-15") {
+		t.Fatal("inclusive end")
+	}
+	if ActiveAt("2024-03-01", "2025-07-15", "2025-07-16") {
+		t.Fatal("day after end")
+	}
+	if ActiveAt("2024-03-01", "2025-07-15", "2024-02-28") {
+		t.Fatal("day before start")
+	}
+}
+
+func TestActiveAtEmptyIntervalAlwaysTrue(t *testing.T) {
+	// legacy leafs without intervals stay visible for any as-of
+	if !ActiveAt("", "", "2025-01-01") {
+		t.Fatal("empty interval must remain active")
+	}
+	if !ActiveAt("", "", "") {
+		t.Fatal("no as-of means all active")
+	}
+}
+
+func TestActiveAtEmptyAsOfKeepsAll(t *testing.T) {
+	if !ActiveAt("2099-01-01", "2099-12-31", "") {
+		t.Fatal("empty as-of must not filter")
+	}
+}
+
+func TestAsOfPickXNotY(t *testing.T) {
+	// Acceptance from #36: as of 2025-01-01 → X, not Y
+	xFrom, xTo := "2024-03-01", "2025-07-15"
+	yFrom, yTo := "2025-07-16", ""
+	asOf := "2025-01-01"
+	if !ActiveAt(xFrom, xTo, asOf) {
+		t.Fatal("X must be active as of 2025-01-01")
+	}
+	if ActiveAt(yFrom, yTo, asOf) {
+		t.Fatal("Y must be inactive as of 2025-01-01")
+	}
+}
+
+func TestNormalizeDayTrimsTime(t *testing.T) {
+	if NormalizeDay("2025-01-01T12:00:00Z") != "2025-01-01" {
+		t.Fatalf("got %q", NormalizeDay("2025-01-01T12:00:00Z"))
+	}
+	if NormalizeDay("2025-01-01") != "2025-01-01" {
+		t.Fatalf("got %q", NormalizeDay("2025-01-01"))
+	}
+}

internal/gitlog/args.go

@@ -0,0 +1,51 @@
+package gitlog
+
+import (
+	"fmt"
+	"time"
+
+	"github.com/eSlider/2dph/internal/cli"
+	"github.com/integrii/flaggy"
+)
+
+type CLI struct {
+	Repo, Root, Since string
+	Limit             int
+	JSONOut           bool
+}
+
+func Parser() *flaggy.Parser {
+	c := CLI{}
+	return Bind(&c)
+}
+
+func Bind(c *CLI) *flaggy.Parser {
+	p := cli.New("git-import")
+	p.Description = "go-git history → commit leafs"
+	p.Bool(&c.JSONOut, "", "json", "JSON output")
+	p.Int(&c.Limit, "", "limit", "max commits (0 = all)")
+	p.String(&c.Since, "", "since", "RFC3339 or YYYY-MM-DD")
+	p.String(&c.Root, "", "root", "scan dir for git repos")
+	p.AddPositionalValue(&c.Repo, "repo", 1, false, "git repo path")
+	return p
+}
+
+func ParseArgs(args []string) (CLI, error) {
+	var c CLI
+	if err := cli.Parse(Bind(&c), args); err != nil {
+		return c, err
+	}
+	return c, nil
+}
+
+func ParseSince(s string) (time.Time, error) {
+	if s == "" {
+		return time.Time{}, nil
+	}
+	for _, layout := range []string{time.RFC3339, "2006-01-02"} {
+		if t, err := time.Parse(layout, s); err == nil {
+			return t, nil
+		}
+	}
+	return time.Time{}, fmt.Errorf("cannot parse --since %q", s)
+}

internal/httpapi/mcp.go

@@ -105,11 +105,18 @@ func (s *Server) mcpCall(r *http.Request, params json.RawMessage) (any, error) {
 		if limit < 1 || limit > 100 {
 			return mcpText(`{"error":"n must be int 1..100"}`, true), nil
 		}
+		asOf := ""
+		if raw, ok := p.Arguments["as_of"]; ok {
+			asOf = strings.TrimSpace(fmt.Sprint(raw))
+			if asOf == "<nil>" {
+				asOf = ""
+			}
+		}
 		if !s.tryAcquire(r) {
 			return nil, fmt.Errorf("cancelled")
 		}
 		defer s.release()
-		body, err = s.api.Search(r.Context(), q, limit)
+		body, err = s.api.Search(r.Context(), q, limit, asOf)
 	case "get":
 		id := strings.TrimSpace(fmt.Sprint(p.Arguments["id"]))
 		if id == "" || id == "<nil>" {

internal/httpapi/server.go

@@ -24,7 +24,7 @@ import (
 
 // API is the in-process brain surface. Production serve.go wires internal/brain.
 type API interface {
-	Search(ctx context.Context, query string, limit int) ([]byte, error)
+	Search(ctx context.Context, query string, limit int, asOf string) ([]byte, error)
 	Get(ctx context.Context, id string, body bool) ([]byte, error)
 	Stats(ctx context.Context) ([]byte, error)
 	Audit(ctx context.Context) ([]byte, error)
@@ -85,11 +85,12 @@ func (s *Server) handleSearch(w http.ResponseWriter, r *http.Request) {
 		}
 		limit = n
 	}
+	asOf := strings.TrimSpace(r.URL.Query().Get("as_of"))
 	if !s.acquire(w, r) {
 		return
 	}
 	defer s.release()
-	body, err := s.api.Search(r.Context(), q, limit)
+	body, err := s.api.Search(r.Context(), q, limit, asOf)
 	writeAPI(w, body, err)
 }
 
@@ -187,13 +188,18 @@ type ExecSearcher struct {
 	Timeout time.Duration
 }
 
-func (b ExecSearcher) Search(ctx context.Context, query string, limit int) ([]byte, error) {
+func (b ExecSearcher) Search(ctx context.Context, query string, limit int, asOf string) ([]byte, error) {
 	if b.Timeout == 0 {
 		b.Timeout = 60 * time.Second
 	}
 	ctx, cancel := context.WithTimeout(ctx, b.Timeout)
 	defer cancel()
-	cmd := exec.CommandContext(ctx, b.CmdPath, "--json", "-n", strconv.Itoa(limit), query)
+	args := []string{"--json", "-n", strconv.Itoa(limit)}
+	if asOf != "" {
+		args = append(args, "--as-of", asOf)
+	}
+	args = append(args, query)
+	cmd := exec.CommandContext(ctx, b.CmdPath, args...)
 	out, err := cmd.Output()
 	if err != nil {
 		var exitErr *exec.ExitError

internal/httpapi/server_test.go

@@ -21,10 +21,10 @@ type fakeSearcher struct {
 	calls    int
 	active   atomic.Int32
 	maxSeen  atomic.Int32
-	callback func(q string, limit int) ([]byte, error)
+	callback func(q string, limit int, asOf string) ([]byte, error)
 }
 
-func (f *fakeSearcher) Search(ctx context.Context, query string, limit int) ([]byte, error) {
+func (f *fakeSearcher) Search(ctx context.Context, query string, limit int, asOf string) ([]byte, error) {
 	f.mu.Lock()
 	f.calls++
 	f.mu.Unlock()
@@ -44,7 +44,7 @@ func (f *fakeSearcher) Search(ctx context.Context, query string, limit int) ([]b
 		}
 	}
 	if f.callback != nil {
-		return f.callback(query, limit)
+		return f.callback(query, limit, asOf)
 	}
 	return []byte(`{"query":"` + query + `","count":0,"results":[]}`), nil
 }
@@ -109,7 +109,7 @@ func TestSearchMissingQuery(t *testing.T) {
 }
 
 func TestSearchReturnsSearcherResult(t *testing.T) {
-	fs := &fakeSearcher{callback: func(q string, limit int) ([]byte, error) {
+	fs := &fakeSearcher{callback: func(q string, limit int, asOf string) ([]byte, error) {
 		return []byte(`{"query":"` + q + `","count":1,"results":[{"id":"x"}]}`), nil
 	}}
 	h := NewServer(fs, 1)
@@ -168,7 +168,7 @@ func TestSearchRejectsBadLimit(t *testing.T) {
 }
 
 func TestGetLeaf(t *testing.T) {
-	fs := &fakeSearcher{callback: func(q string, limit int) ([]byte, error) {
+	fs := &fakeSearcher{callback: func(q string, limit int, asOf string) ([]byte, error) {
 		return []byte(`{}`), nil
 	}}
 	h := NewServer(fs, 1)

internal/httpapi/spec.go

@@ -35,6 +35,7 @@ var Ops = []Op{
 		Params: []Param{
 			{Name: "q", In: "query", Type: "string", Description: "search query", Required: true},
 			{Name: "n", In: "query", Type: "integer", Description: "hit limit 1..100 (default 10)"},
+			{Name: "as_of", In: "query", Type: "string", Description: "YYYY-MM-DD; keep facts active on that day (D24)"},
 		},
 	},
 	{
@@ -54,6 +55,8 @@ var Ops = []Op{
 			{Name: "text", In: "query", Type: "string", Description: "leaf text (omit for CLI hint)"},
 			{Name: "root", In: "query", Type: "string", Description: "facts or info (default info)"},
 			{Name: "source", In: "query", Type: "string", Description: "evidence pointer; facts need two sources"},
+			{Name: "valid_from", In: "query", Type: "string", Description: "fact interval start YYYY-MM-DD (D24)"},
+			{Name: "valid_to", In: "query", Type: "string", Description: "fact interval end YYYY-MM-DD inclusive (D24)"},
 		},
 	},
 	{Path: PathOpenAPI, Method: "get", ID: "openapi", Summary: "OpenAPI 3 document for this server"},

internal/mdleaves/args.go

@@ -0,0 +1,41 @@
+package mdleaves
+
+import (
+	"github.com/eSlider/2dph/internal/cli"
+	"github.com/integrii/flaggy"
+)
+
+type CLI struct {
+	Root    string
+	Files   string
+	JSONOut bool
+}
+
+func Parser() *flaggy.Parser {
+	c := CLI{Root: "."}
+	return Bind(&c)
+}
+
+func Bind(c *CLI) *flaggy.Parser {
+	if c.Root == "" {
+		c.Root = "."
+	}
+	p := cli.New("markdown-import")
+	p.Description = "split markdown H2 leafs"
+	p.Bool(&c.JSONOut, "", "json", "JSON output")
+	p.String(&c.Files, "", "files", "comma-separated paths")
+	p.AddPositionalValue(&c.Root, "dir", 1, false, "markdown root")
+	return p
+}
+
+func ParseArgs(args []string) (CLI, error) {
+	c := CLI{Root: "."}
+	p := Bind(&c)
+	if err := cli.Parse(p, args); err != nil {
+		return c, err
+	}
+	if extra := cli.Query("", p.TrailingArguments); extra != "" && c.Root == "." {
+		c.Root = extra
+	}
+	return c, nil
+}

internal/ocr/args.go

@@ -0,0 +1,35 @@
+package ocr
+
+import (
+	"fmt"
+
+	"github.com/eSlider/2dph/internal/cli"
+	"github.com/integrii/flaggy"
+)
+
+type CLI struct {
+	Path string
+}
+
+func Parser() *flaggy.Parser {
+	c := CLI{}
+	return Bind(&c)
+}
+
+func Bind(c *CLI) *flaggy.Parser {
+	p := cli.New("mail-ocr")
+	p.Description = "tesseract eng+deu on image or scanned PDF"
+	p.AddPositionalValue(&c.Path, "file", 1, false, "image or pdf")
+	return p
+}
+
+func ParseArgs(args []string) (CLI, error) {
+	var c CLI
+	if err := cli.Parse(Bind(&c), args); err != nil {
+		return c, err
+	}
+	if c.Path == "" {
+		return c, fmt.Errorf("usage: bin/mail/ocr.go <image|pdf>")
+	}
+	return c, nil
+}

internal/reasoner/args.go

@@ -0,0 +1,47 @@
+package reasoner
+
+import (
+	"os"
+
+	"github.com/eSlider/2dph/internal/cli"
+	"github.com/integrii/flaggy"
+)
+
+type CLI struct {
+	Base    string
+	Model   string
+	Device  string
+	JSONOut bool
+}
+
+func Parser() *flaggy.Parser {
+	c := NewCLI()
+	return Bind(&c)
+}
+
+func NewCLI() CLI {
+	base := os.Getenv("REASONER_BASE_URL")
+	if base == "" {
+		base = "http://127.0.0.1:11435/v1"
+	}
+	model := os.Getenv("REASONER_MODEL")
+	if model == "" {
+		model = OllamaRAM
+	}
+	return CLI{Base: base, Model: model, Device: "cpu"}
+}
+
+func Bind(c *CLI) *flaggy.Parser {
+	p := cli.New("reasoner-bakeoff")
+	p.Description = "CPU tool-call bake-off"
+	p.Bool(&c.JSONOut, "", "json", "JSON output")
+	p.String(&c.Model, "", "model", "Ollama/HF model id")
+	p.String(&c.Base, "", "base-url", "OpenAI-compatible URL")
+	p.String(&c.Device, "", "device", "cpu")
+	return p
+}
+
+func ParseArgs(args []string) (CLI, error) {
+	c := NewCLI()
+	return c, cli.Parse(Bind(&c), args)
+}

internal/websearch/args.go

@@ -0,0 +1,63 @@
+package websearch
+
+import (
+	"fmt"
+
+	"github.com/eSlider/2dph/internal/cli"
+	"github.com/integrii/flaggy"
+)
+
+type CLI struct {
+	Query, Site, Lang, Fresh, Category, Engines string
+	Limit                                       int
+	JSONOut, Refresh, Force                     bool
+	TTL                                         float64
+	Timeout                                     int
+}
+
+func NewCLI() CLI {
+	return CLI{Limit: DefaultLimit, TTL: float64(CacheTTL), Timeout: 25}
+}
+
+func Parser() *flaggy.Parser {
+	c := NewCLI()
+	return Bind(&c)
+}
+
+func Bind(c *CLI) *flaggy.Parser {
+	p := cli.New("web-search")
+	p.Description = "SearXNG second source (throttled ≠ absence)"
+	p.Bool(&c.JSONOut, "", "json", "JSON output")
+	p.Bool(&c.Refresh, "", "refresh", "bypass cache")
+	p.Bool(&c.Force, "", "force", "allow PII in query")
+	p.Int(&c.Limit, "n", "limit", "max hits")
+	p.String(&c.Site, "", "site", "restrict to host")
+	p.String(&c.Lang, "", "lang", "language")
+	p.String(&c.Fresh, "", "fresh", "day|week|month|year")
+	p.String(&c.Category, "", "category", "searx category")
+	p.String(&c.Engines, "", "engines", "engine list")
+	p.Float64(&c.TTL, "", "ttl", "cache ttl seconds")
+	p.Int(&c.Timeout, "", "timeout", "http timeout seconds")
+	return p
+}
+
+func ParseArgs(args []string) (CLI, error) {
+	c := NewCLI()
+	p := Bind(&c)
+	var q string
+	p.AddPositionalValue(&q, "query", 1, false, "search query")
+	if err := cli.Parse(p, args); err != nil {
+		return c, err
+	}
+	c.Query = cli.Query(q, p.TrailingArguments)
+	if c.Query == "" {
+		return c, fmt.Errorf("query required")
+	}
+	if c.Limit < 0 {
+		return c, fmt.Errorf("--limit must be a non-negative integer")
+	}
+	if c.Timeout <= 0 {
+		return c, fmt.Errorf("--timeout must be a positive integer")
+	}
+	return c, nil
+}

skills/brain/SKILL.md

@@ -26,6 +26,7 @@ second independent source when local roots cannot confirm. An answer is
 bin/brain/search.go "Matrix federation"                # pointers + snippets, YAML
 bin/brain/search.go "onlyoffice postgres" --root facts # restrict to confirmed
 bin/brain/search.go "where is cs-lexicon" --json | yq '.[].ref'
+bin/brain/search.go "who works where" --as-of 2025-01-01  # D24 intervals
 bin/brain/add.go --text T --root facts --source "a.md x b.md"
 bin/brain/get.go <id> --body                           # full chunk only when needed
 bin/brain/stats.go                                     # index health
@@ -34,6 +35,8 @@ bin/brain/eval.go                                      # recall@5 >= 0.95 gate (
 
 `bin/kb/search` is a deprecated wrapper. `--hop N` walks
 `FROM_FILE` / `HAS_VERSION` / `AUTHORED` from each hit (1=File, 3=Person).
+`--as-of YYYY-MM-DD` keeps leafs whose `valid_from`/`valid_to` cover that day
+(empty interval = always; not D16 source staleness).
 
 ## Rules
 
@@ -45,6 +48,8 @@ bin/brain/eval.go                                      # recall@5 >= 0.95 gate (
   are not evidence of absence. `--root facts|info` and `--no-web` skip the web.
 - If recall looks wrong, run `bin/brain/eval.go`; it gates control questions and
   should stay at or above 95% recall@5.
+- Contradictions (≥2 yes vs ≥2 no) stay `(not confirmed)` until
+  `bin/facts/audit contradict` fires `temporal_freshness` or `authority_pairing`.
 - Agents: `GET /openapi.json` and `POST /mcp` on `bin/brain/serve.go` (same
   handlers; tool names match paths `search`/`get`/`stats`/`audit`). Generated
   list: [tools.md](tools.md).

skills/picoclaw/SKILL.md

@@ -12,6 +12,12 @@ PicoClaw speaks MCP at `POST /mcp` on `bin/brain/serve.go`. Compose profile
 `picoclaw` runs the official `sipeed/picoclaw` gateway plus `brain-mcp`
 (see [docs/picoclaw.md](../../docs/picoclaw.md)).
 
+```bash
+bin/stack/start-assistant     # brain + qwen3.5:9b + gateway + picoclaw agent
+bin/stack/status
+bin/stack/stop
+```
+
 ## Tool order (before a factual reply)
 
 1. **`search`** — facts root first, then info. The `web` block is a second