docs(tools): placement doctrine for Tools.md, plus the pointer rot the old model left behind

LifeOS/install/LIFEOS/DOCUMENTATION/Delegation/DelegationSystem.md

@@ -198,6 +198,87 @@ Three primitives for non-blocking work. Pick the right one:
 - Use `TaskStop` to cancel a monitor early
 - Selective filters only — never pipe raw logs. Monitors producing too many events get auto-stopped.
 
+### Recipe: Agent Watchdog (auto-triggered by hook)
+
+The Pulse agent-guard hook automatically injects a watchdog reminder when background agents are spawned. The watchdog monitors tool-activity.jsonl for silence while agents are active:
+
+```bash
+Monitor({
+  description: "Agent watchdog",
+  persistent: true,
+  timeout_ms: 3600000,
+  command: "bun $HOME/.claude/LIFEOS/TOOLS/AgentWatchdog.ts"
+})
+```
+
+Alerts when no tool calls detected for 90 seconds with active agents. Rate-limited to one alert per 60 seconds. Runs for the session lifetime — covers all background agents.
+
+### Recipe: Deploy Monitoring
+
+Watch a Cloudflare Pages or Workers deploy for completion or errors:
+
+```bash
+# Monitor wrangler deploy output (run deploy with Bash(run_in_background), then tail its output)
+Monitor({
+  description: "Cloudflare deploy status",
+  persistent: false,
+  timeout_ms: 300000,
+  command: "tail -f /tmp/deploy.log | grep --line-buffered -E '(Published|Error|Failed|SUCCESS)'"
+})
+```
+
+### Recipe: Pulse Log Tailing
+
+Watch Pulse daemon logs for errors during a debugging session:
+
+```bash
+Monitor({
+  description: "Pulse error watcher",
+  persistent: true,
+  timeout_ms: 300000,
+  command: "tail -f ~/.claude/Pulse/logs/pulse-stdout.log | grep --line-buffered -i -E '(error|fatal|crash|unhandled)'"
+})
+```
+
+### Recipe: Build/Test Watching
+
+Monitor a long test suite and get notified on failures:
+
+```bash
+Monitor({
+  description: "Test failure watcher",
+  persistent: false,
+  timeout_ms: 600000,
+  command: "tail -f /tmp/test-output.log | grep --line-buffered -E '(FAIL|ERROR|✗|AssertionError)'"
+})
+```
+
+### Recipe: PR/CI Status Monitoring
+
+Poll GitHub for CI status changes on a PR:
+
+```bash
+Monitor({
+  description: "CI status for PR #42",
+  persistent: true,
+  timeout_ms: 3600000,
+  command: "last_status=''; while true; do status=$(gh pr checks 42 --json state --jq '.[].state' 2>/dev/null | sort -u | tr '\\n' ','); if [ \"$status\" != \"$last_status\" ]; then echo \"CI: $status\"; last_status=\"$status\"; fi; sleep 30; done"
+})
+```
+
+### Recipe: Security Scan Watching
+
+Tail security scan results for critical findings:
+
+```bash
+Monitor({
+  description: "Security scan critical findings",
+  persistent: false,
+  timeout_ms: 600000,
+  command: "tail -f /tmp/security-scan.log | grep --line-buffered -i 'CRITICAL'"
+})
+```
+
 ---
 
 ## Knowledge Archive Access

LifeOS/install/LIFEOS/DOCUMENTATION/LifeosSystemArchitecture.md

@@ -587,7 +587,7 @@ System file inventory by pipeline. When you modify a file, trace its pipeline to
 | **Security** | `LIFEOS/LIFEOS_SYSTEM_PROMPT.md` § Security Protocol (constitutional rule), `settings.json` `permissions.deny` (native harness denylist), `hooks/Safety.hook.ts` (consolidated PermissionRequest + PostToolUse on WebFetch \| WebSearch), `hooks/lib/safety-classifier.ts` (shape catalog + shell-aware classifier with single-quote pre-pass) |
 | **Algorithm** | `Algorithm/LATEST` → `Algorithm/v{VERSION}.md` (currently v8.4.0 — the claims restructure: Loop preamble + 15 teeth-annotated done-claims + AlgorithmNudge event layer (unified 2026-07-11: run-scoped + always-on skill-routing/late-ISA/spend; depth-directive row added v8.4.0, 2026-07-12); capabilities.md removed at v7, the system-prompt skill list is the sole capability inventory), `Algorithm/mode-detection.md`, `hooks/ISASync.hook.ts` → `MEMORY/WORK/{slug}/ISA.md`, `skills/ISA/` (canonical Scaffold/Append/Reconcile workflows); **work registry event-sourced (2026-06-10):** all `work.json` writes go through `isa-utils.writeRegistry` → field-level diff events appended to `MEMORY/STATE/work-events.jsonl` (`hooks/lib/work-events.ts`) → locked fold to the derived `work.json` snapshot (offset-stamped, 1MB compaction); `readRegistry` serves snapshot+suffix live views; Pulse SSE triggers off `fs.watch` on STATE with the 100ms poll as fallback; model choice is per-dispatch judgment (2026-07-11 baseline); `EFFORT_MODEL` in `LIFEOS/TOOLS/models.ts` remains the `Inference.ts --level` dial |
 | **Memory** | `hooks/WorkCompletionLearning.hook.ts`, `hooks/SatisfactionCapture.hook.ts` (RelationshipMemory hook deleted 7.0.0 — dead code), `Tools/KnowledgeHarvester.ts` → `MEMORY/KNOWLEDGE/`, `MEMORY/LEARNING/`; `Tools/SessionHarvester.ts --mine` → `KNOWLEDGE/_harvest-queue/`; `Tools/MemoryRetriever.ts` (BM25 retrieval over typed-item corpus including `_MEMORY.md` hot-layer files), `Tools/KnowledgeGraph.ts` (graph navigation) — read-only. **Autonomic loop (2026-05):** `hooks/MemoryTurnStart.hook.ts` (UserPromptSubmit) + `hooks/MemoryReviewFire.hook.ts` (Stop; cadence merged 2026-07-11) drive `Tools/MemoryReviewer.ts` on cadence (turn≥8 ∧ minutes≥30 ∧ idle≥2). Reviewer emits typed items routed by `Tools/MemorySystem.ts` (single `add(item)` API) over the `Tools/MemoryTypes.ts` registry; `Tools/MutationTier.ts` gates by tier A/B/C/D. Tier-C proposals enqueue to `MEMORY/OBSERVABILITY/pending-proposals.jsonl`; `PULSE/lib/telegram-proposals.ts` + `PULSE/modules/telegram.ts` surface them as `yes/no/edit #id` Telegram replies. `Tools/MemoryStatus.ts` is the read-only `kai status` CLI. **kb-v3 knowledge schema (2026-07-05):** `Tools/KnowledgeSchema.ts` is the pure-data SoT for the KNOWLEDGE archive object-schema (`person\|company\|idea\|blog\|research` — distinct from the write-registry above) + body-safe parse/normalize/validate; `Tools/KnowledgeLint.ts` validates conformance (envelope % vs per-type completeness); `Tools/MigrateKnowledge.ts` migrated ~4,400 notes onto it (body-byte-preserving, idempotent, dry-run default); `Tools/KnowledgeQuery.ts` (`kb query`) filters/sorts on the now-consistent typed fields; `Tools/GenerateKnowledgeSchemaDoc.ts` regenerates `MEMORY/KNOWLEDGE/_schema.md` from the schema; `MemorySystem.renderInitialNote` emits the kb-v3 envelope so new autonomic notes are born conformant. |
-| **Router** (RETIRED 2026-07-11) | Classify → route-effort stages retired 2026-07-11 with the mode/tier abolition (`TheRouter.hook.ts` deleted; MINIMAL/NATIVE/ALGORITHM + E1–E5 gone, no successor classifier). **Surviving model routing:** `LIFEOS/TOOLS/models.ts` `EFFORT_MODEL` maps level→model (max→fable / high→opus / medium→sonnet / low→haiku; `LEVEL_TO_HARNESS_EFFORT`; cross-vendor pins; egress-class ceilings) → **dispatch** via `model` param on `Agent()` / `Workflow agent()`, injected by `hooks/AgentInvocation.hook.ts` on unspecified dispatches. `LIFEOS/TOOLS/Inference.ts` applies model selection to utility inference, and is the genuine `max`/Fable carrier (subprocess spawns `claude --model claude-fable-5`; Agent `model:fable` dispatch downgrades to Opus). It verifies the executed model against the JSON envelope's `modelUsage` and logs downgrades to `MEMORY/OBSERVABILITY/model-verification.jsonl` (v6.29.0 — reports what RAN, not what was requested). Full doc (history only): `LIFEOS/DOCUMENTATION/Router/RouterSystem.md` |
+| **Router** (RETIRED 2026-07-11) | Classify → route-effort stages retired 2026-07-11 with the mode/tier abolition (`TheRouter.hook.ts` deleted; MINIMAL/NATIVE/ALGORITHM + E1–E5 gone, no successor classifier). **Surviving model routing:** `LIFEOS/TOOLS/models.ts` `EFFORT_MODEL` maps level→model (max→fable / high→opus / medium→sonnet / low→haiku; `LEVEL_TO_HARNESS_EFFORT`; cross-vendor pins; egress-class ceilings) → **dispatch** via `model` param on `Agent()` / `Workflow agent()`, injected by `hooks/AgentInvocation.hook.ts` on unspecified dispatches. `LIFEOS/TOOLS/Inference.ts` applies model selection to utility inference and is the genuine `max`/Fable carrier with executed-model verification — mechanism: `LIFEOS/DOCUMENTATION/Router/RouterSystem.md` § Carrier + verification (also the full Router doc, history only). |
 | **Hooks** | `hooks/*.hook.ts`, `hooks/handlers/*.ts`, `hooks/lib/*.ts`, `settings.json` |
 | **Observability** | `hooks/EventLogger.hook.ts` (consolidated 2026-07-11 — absorbed ToolActivityTracker, ToolFailureTracker, SkillExecutionLog, ConfigAudit, StopFailureHandler; appends directly via `fs.appendFileSync`) → `MEMORY/OBSERVABILITY/*.jsonl` |
 | **Pulse** | `Pulse/pulse.ts` (port 31337), `Pulse/modules/{observability,hooks,wiki,imessage,telegram,user-index,da,work,bunker}.ts`, `Pulse/PULSE.toml`, `Pulse/Observability/src/`, `Pulse/Assistant/module.ts` |

LifeOS/install/LIFEOS/DOCUMENTATION/Notifications/NotificationSystem.md

@@ -124,6 +124,35 @@ curl -s -X POST http://localhost:31337/notify \
 
 ---
 
+## Long-form Narration
+
+The same `/notify` endpoint reads longer text aloud — "read this to me", "narrate this", "speak this", "perform this". It is a Pulse route, not a callable tool, so it lives here rather than in the tool index.
+
+```bash
+# Single narration segment
+curl -X POST http://localhost:31337/notify \
+  -H "Content-Type: application/json" \
+  -d '{
+    "message": "Your text here",
+    "voice_id": "$ELEVENLABS_VOICE_ID",
+    "title": "Voice Narrative"
+  }'
+
+# Pause between segments
+sleep 2
+```
+
+**Narration settings:**
+- **Voice ID:** `ELEVENLABS_VOICE_ID` environment variable
+- **Stability:** 0.55 (natural variation in storytelling)
+- **Similarity Boost:** 0.85 (maintains authentic sound)
+- **Max segment:** 450 characters — split longer text
+- **Pause between:** 2 seconds, for storytelling flow
+
+Pulse must be running; the voice handler lives at `~/.claude/LIFEOS/PULSE/VoiceServer/voice.ts` on port 31337, and calls ElevenLabs under the hood.
+
+---
+
 ## Copy-Paste Templates
 
 ### Template A: Skills WITH Workflows