From 06eaf5820a84ca64b162cb9eb51ecbc917a8fc99 Mon Sep 17 00:00:00 2001
From: richsak <richie@richiesakhon.com>
Date: Sun, 26 Apr 2026 15:54:59 -0700
Subject: [PATCH 1/2] feat: add webster-weekly-council library skill

Library-shape conversion of prompts/second-wbs-session.md so operators run
the weekly council pass via /webster-weekly-council. SKILL.md is a slim
phase index; nine references/ files hold per-phase bash blocks loaded on
demand; two skill-local scripts/ extract the reusable polling helper and
the planner-JSON parser. The 662-line single-page prompt remains intact
as the locked source-of-truth runbook.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---
 skills/webster-weekly-council/SKILL.md        |  66 +++++++++++
 .../references/branch.md                      |  25 ++++
 .../references/failure-modes.md               |  29 +++++
 .../references/fan-out.md                     |  86 ++++++++++++++
 .../references/genealogy.md                   |  76 ++++++++++++
 .../references/planner.md                     | 108 +++++++++++++++++
 .../references/preflight.md                   |  98 ++++++++++++++++
 .../references/publish.md                     | 109 ++++++++++++++++++
 .../references/redesign.md                    |  63 ++++++++++
 .../references/seed-history.md                |  74 ++++++++++++
 .../references/verify.md                      |  63 ++++++++++
 .../scripts/extract-planner-json.py           |  26 +++++
 .../scripts/run-agent-session.sh              |  85 ++++++++++++++
 13 files changed, 908 insertions(+)
 create mode 100644 skills/webster-weekly-council/SKILL.md
 create mode 100644 skills/webster-weekly-council/references/branch.md
 create mode 100644 skills/webster-weekly-council/references/failure-modes.md
 create mode 100644 skills/webster-weekly-council/references/fan-out.md
 create mode 100644 skills/webster-weekly-council/references/genealogy.md
 create mode 100644 skills/webster-weekly-council/references/planner.md
 create mode 100644 skills/webster-weekly-council/references/preflight.md
 create mode 100644 skills/webster-weekly-council/references/publish.md
 create mode 100644 skills/webster-weekly-council/references/redesign.md
 create mode 100644 skills/webster-weekly-council/references/seed-history.md
 create mode 100644 skills/webster-weekly-council/references/verify.md
 create mode 100755 skills/webster-weekly-council/scripts/extract-planner-json.py
 create mode 100755 skills/webster-weekly-council/scripts/run-agent-session.sh

diff --git a/skills/webster-weekly-council/SKILL.md b/skills/webster-weekly-council/SKILL.md
new file mode 100644
index 0000000..b016e3e
--- /dev/null
+++ b/skills/webster-weekly-council/SKILL.md
@@ -0,0 +1,66 @@
+---
+name: webster-weekly-council
+description: Run one full Webster weekly landing-page council pass — pre-flight, planner, 6 parallel managed-agent critics, genealogy gap-detection, redesigner synthesis, and draft PR. Use when the operator runs /webster-weekly-council or asks for the weekly Webster council run. Replaces `wbs @prompts/second-wbs-session.md` as the primary operator path; the prompt remains the readable single-page fallback runbook.
+---
+
+# Webster Weekly Council (Library)
+
+One full council pass: planner → 6 parallel critics → genealogy → redesigner → draft PR. **30–50 min wall-clock. ~$0.16–0.25 in API tokens per run.**
+
+> **Override default Operating Loop.** This is a council-run session. Execute the phases below end-to-end. Do NOT call `forge isolation list`, do NOT scan `FEATURES.md` for `todo` rows, do NOT enter a feature-implementation loop. The phases are sequential; do not skip ahead.
+
+## Phase index — load on demand
+
+Read `references/preflight.md` first. After that, load only the reference for the phase you're executing. Do not load all references up front.
+
+| #   | Phase                                                | Wall-clock | Reference                     |
+| --- | ---------------------------------------------------- | ---------- | ----------------------------- |
+| 0   | Pre-flight + session constants                       | <1 min     | `references/preflight.md`     |
+| 1   | Seed 10-week mock analytics history (idempotent)     | 2 min      | `references/seed-history.md`  |
+| 2   | Prepare shared `council/<date>` branch               | 1 min      | `references/branch.md`        |
+| 3   | Run planner (Opus 4.7) → `plan.md` + memory event    | 3–5 min    | `references/planner.md`       |
+| 4   | Fan-out 6 parallel managed-agent sessions            | 15–20 min  | `references/fan-out.md`       |
+| 5   | Verify findings (gate before redesigner)             | 2 min      | `references/verify.md`        |
+| 5.5 | Critic Genealogy gap-detection (fail-open)           | 0–10 min   | `references/genealogy.md`     |
+| 6   | Redesigner session → `proposal.md` + `decision.json` | 5–10 min   | `references/redesign.md`      |
+| 7+8 | Open draft PR + write checkpoint                     | 1 min      | `references/publish.md`       |
+| —   | If a step fails                                      | —          | `references/failure-modes.md` |
+
+## Always-true invariants
+
+These hold across every phase. Do not break them.
+
+1. **`ANTHROPIC_API_KEY` is fetched from macOS keychain into a scoped bash variable.** Never read it from the operator's shell env — that would bill `claude -p` and any Forge call against API credits instead of the Max sub.
+2. **Registration artifacts must exist before fan-out.** `environments/webster-council-env.id`, `context/monitor/id.txt`, `context/redesigner/id.txt`, and the 5 critic id files. If missing, run `prompts/first-wbs-session.md` first.
+3. **All writes go to `council/$WEEK_DATE`.** Local working tree and remote committers (critics + redesigner via GitHub MCP) all target the same branch. The orchestrator only creates the branch locally; critics' system prompts handle remote create-or-skip.
+4. **Phase 3 (planner) is fail-closed.** If planner errors, halt before fan-out. Redesigner has no findings to synthesize without it.
+5. **Phase 5.5 (genealogy) is fail-open.** Errors log and continue. Redesigner runs against the 5 original critic findings if genealogy fails.
+6. **Phase 5 (verify) is the redesigner gate.** If <3 critics produced non-stub findings, abort before spending redesigner tokens.
+7. **Findings are sovereign over planner direction_hint.** Critics report what they find regardless of plan; redesigner uses the hint as a weighting input but CRITICAL/HIGH critic evidence overrides.
+8. **Don't merge the PR.** It opens as a draft. Human review = approval. The operator's job ends at PR-open + checkpoint.
+
+## Helper scripts ("unlocked tools")
+
+These are reusable across phases. Each is a self-contained CLI; pass args from the bash blocks in references.
+
+- **`scripts/run-agent-session.sh`** — create a managed-agent session, send a `user.message`, poll until idle/completed/timeout. Used 7× across phases 4 and 6 (monitor + 5 critics + redesigner). Exit code: `0` ok, `1` failed, `2` timeout.
+- **`scripts/extract-planner-json.py`** — pull the JSON block out of `history/<date>/plan.md` for the optional `new_critic_request` extraction in phase 3.
+
+Both scripts are skill-local under `skills/webster-weekly-council/scripts/`. Repo-level helpers (`scripts/planner-invoke.ts`, `scripts/planner-context.ts`, `scripts/critic-genealogy.ts`, `scripts/validate-findings.ts`) are unchanged and still called by reference bash blocks.
+
+## Production invariant — `prompts/second-wbs-session.md`
+
+The 662-line bash-in-markdown prompt is the immutable production runbook. This skill is the operator surface; the prompt is the source-of-truth runbook. They produce the same artifacts.
+
+- The prompt is locked by `scripts/__tests__/sim-council.test.ts:75` — `git diff prompts/second-wbs-session.md` must be empty.
+- Nicolette's live council on `main` runs the prompt via `wbs @prompts/second-wbs-session.md`.
+- If the skill drifts from the prompt, **fix the skill** — never the prompt.
+
+## Quick start
+
+1. Load `references/preflight.md` → run pre-flight + session constants.
+2. Load each phase reference in order (1 → 2 → 3 → 4 → 5 → 5.5 → 6 → 7+8).
+3. If any phase fails, load `references/failure-modes.md` and follow the named recovery.
+4. Final state: draft PR open, checkpoint committed under `.claude/checkpoints/`.
+
+If a phase genuinely cannot be expressed as a skill instruction, surface `[STUCK]` per `AGENTS.md` and stop. Visible struggle > invisible corner-cutting.
diff --git a/skills/webster-weekly-council/references/branch.md b/skills/webster-weekly-council/references/branch.md
new file mode 100644
index 0000000..2cb582d
--- /dev/null
+++ b/skills/webster-weekly-council/references/branch.md
@@ -0,0 +1,25 @@
+# Phase 2 — Prepare shared `council/<date>` branch
+
+**~1 min.**
+
+All 6 parallel workers + the redesigner commit to `$BRANCH` via GitHub MCP. The critics' system prompts already create-or-skip the branch via `create_branch` MCP (422-on-exists is treated as success), so this local setup is about keeping your working tree consistent — not about creating the branch remotely.
+
+```bash
+git checkout main
+git pull
+git fetch origin
+if git ls-remote --heads origin "$BRANCH" | grep -q "$BRANCH"; then
+  echo "Branch $BRANCH already exists on origin — reusing"
+  git checkout -B "$BRANCH" "origin/$BRANCH"
+else
+  git checkout -B "$BRANCH" main
+  git push -u origin "$BRANCH"
+fi
+```
+
+## Why both branches are valid
+
+- **Branch already exists**: a prior attempt this week (failed mid-run, or genealogy/redesigner re-run). Reuse so the new orchestrator-side commits land alongside the existing critic findings.
+- **Branch missing**: fresh weekly run. Create from `main` and push so MCP-side critic commits don't race local creation.
+
+The local working tree must match `origin/$BRANCH` before phase 3, otherwise the planner commit in phase 3 would diverge from the branch the critics will read.
diff --git a/skills/webster-weekly-council/references/failure-modes.md b/skills/webster-weekly-council/references/failure-modes.md
new file mode 100644
index 0000000..27fcc24
--- /dev/null
+++ b/skills/webster-weekly-council/references/failure-modes.md
@@ -0,0 +1,29 @@
+# Failure modes — recovery guide
+
+The skill is idempotent end-to-end. Any phase can be re-run after fixing the named cause.
+
+| Failure                                                    | Recovery                                                                                                                                                                                                                               |
+| ---------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| **Pre-flight abort** (key/keychain/registration artifacts) | Fix the named prerequisite. Re-run the whole skill — preflight is idempotent and every later phase checks for its own outputs before running.                                                                                          |
+| **Planner fails (phase 3)**                                | Run halts before fan-out. Read `tmp/logs/planner.log`, fix the named context/API/JSON problem, re-run from phase 3. Do not continue without `history/$WEEK_DATE/plan.md`.                                                              |
+| **One critic times out (phase 4)**                         | Re-run only that critic: `bash skills/webster-weekly-council/scripts/run-agent-session.sh <label> <id> "$MSG_CRITIC"`. Then re-run phase 5 onward. Don't re-run the whole fan-out.                                                     |
+| **Multiple critics fail (phase 4)**                        | If `${#SUCCEEDED[@]} < 3` in phase 5, abort and investigate. `tmp/logs/<label>.log` shows the last status per agent. API 5xx storms are usually transient — wait 2 min, retry the failed labels.                                       |
+| **Genealogy errors (phase 5.5)**                           | Non-blocking by design. `GENEALOGY_STATUS` captures the failure mode; the pipeline proceeds to the redesigner. To retry standalone: `bun scripts/critic-genealogy.ts --branch "$BRANCH" --week "$WEEK_DATE" --lp-target "$LP_TARGET"`. |
+| **Redesigner fails (phase 6)**                             | Findings are already on `$BRANCH`. Re-run just phase 6. PR can still be opened in phase 7 with findings alone (redesigner output becomes MISSING in the body).                                                                         |
+| **Context budget tight** (autocompact at 20%)              | This skill is designed to run in one pass; checkpoint-before-compact is your safety net. The dispatcher will autocheckpoint, then this skill can be re-run.                                                                            |
+| **Anthropic API 5xx on session create**                    | Retry the single curl. Do not re-run phases 1–2.                                                                                                                                                                                       |
+| **`gh pr create` fails**                                   | The fallback path (no `--json`) usually works. If both fail: `gh auth status`, then create the PR manually with `gh pr create --draft --base main --head "$BRANCH" --title "council: week $WEEK_DATE proposal" --body-file <file>`.    |
+| **Working tree dirty before phase 2**                      | The branch checkout will refuse. `git stash` or commit unrelated work first. Don't `--force` past it.                                                                                                                                  |
+
+## When NOT to retry
+
+- **Pre-flight registration artifact missing**: do NOT seed empty placeholder files. Run `prompts/first-wbs-session.md` (the bootstrap) instead. The artifacts must come from real `POST /v1/agents` registrations.
+- **Findings file is non-empty but malformed**: do NOT hand-edit. Re-run the critic — format drift is a data-quality bug, not a syntax fix.
+
+## When to surface `[STUCK]`
+
+Per `AGENTS.md`: if a path isn't clear, leave a `[STUCK]` prefix in your session output and stop. Don't compose a workaround that hides the difficulty. Examples:
+
+- "Skill instruction conflicts with what `prompts/second-wbs-session.md` says" → `[STUCK]`, name the conflict, stop.
+- "Re-run produced different findings than expected" → `[STUCK]`, surface what you saw vs what you expected.
+- "API 5xx persists across 3 retries" → `[STUCK]`, share the response body, ask whether to wait or escalate.
diff --git a/skills/webster-weekly-council/references/fan-out.md b/skills/webster-weekly-council/references/fan-out.md
new file mode 100644
index 0000000..0a361b0
--- /dev/null
+++ b/skills/webster-weekly-council/references/fan-out.md
@@ -0,0 +1,86 @@
+# Phase 4 — Fan-out 6 parallel managed-agent sessions
+
+**15–20 min. The hot path.** Monitor + 5 critics run in parallel against the same plan and branch.
+
+This phase calls `skills/webster-weekly-council/scripts/run-agent-session.sh` once per agent. The script is the canonical bash for "create session → send `user.message` → poll until idle/completed". Source-of-truth: that script. The bash here only constructs user.messages and orchestrates the 6 parallel jobs.
+
+## User-message templates
+
+```bash
+# Critic message — applies to brand-voice / fh-compliance / seo / conversion / copy
+MSG_CRITIC="BRANCH=$BRANCH
+WEEK_DATE=$WEEK_DATE
+LP_TARGET=$LP_TARGET
+PLAN_PATH=$PLAN_PATH
+
+Planner context (additive only; do not suppress, downgrade, or ignore findings outside this direction):
+$PLAN_TEXT"
+
+# Monitor message — different framing because monitor reports anomalies, not audit findings
+MSG_MONITOR="BRANCH=$BRANCH
+WEEK_DATE=$WEEK_DATE
+PREV_WEEK_DATE=$PREV_WEEK_DATE
+PLAN_PATH=$PLAN_PATH
+
+Planner context (additive only; report anomalies even if they fall outside this direction):
+$PLAN_TEXT"
+```
+
+The "additive only" framing is load-bearing: critics must report what they see, not just what the planner asked about. Findings are sovereign over plan direction.
+
+## Launch all 6 in parallel
+
+```bash
+RUN_AGENT="bash skills/webster-weekly-council/scripts/run-agent-session.sh"
+
+PIDS=()
+$RUN_AGENT monitor          "$MONITOR_ID"         "$MSG_MONITOR"  & PIDS+=($!)
+$RUN_AGENT brand-voice      "$BRAND_VOICE_ID"     "$MSG_CRITIC"   & PIDS+=($!)
+$RUN_AGENT fh-compliance    "$FH_COMPLIANCE_ID"   "$MSG_CRITIC"   & PIDS+=($!)
+$RUN_AGENT seo              "$SEO_ID"             "$MSG_CRITIC"   & PIDS+=($!)
+$RUN_AGENT conversion       "$CONVERSION_ID"      "$MSG_CRITIC"   & PIDS+=($!)
+$RUN_AGENT copy             "$COPY_ID"            "$MSG_CRITIC"   & PIDS+=($!)
+
+# Collect per-agent exit codes (0=ok, 1=failed, 2=timeout)
+declare -A RESULT
+LABELS=(monitor brand-voice fh-compliance seo conversion copy)
+for i in "${!PIDS[@]}"; do
+  wait "${PIDS[$i]}"
+  RESULT["${LABELS[$i]}"]=$?
+done
+
+echo ""
+echo "── Fan-out results ─────────────────────────"
+for L in "${LABELS[@]}"; do
+  case "${RESULT[$L]}" in
+    0) echo "  ✓ $L" ;;
+    1) echo "  ✗ $L — failed (see tmp/logs/$L.log)" ;;
+    2) echo "  ⧗ $L — timed out" ;;
+  esac
+done
+```
+
+## What the script does (read once)
+
+`run-agent-session.sh <label> <agent_id> <user_message>`:
+
+1. `POST /v1/sessions` with `{agent, environment_id, vault_ids, title}` — vault attachment is what gives the session GitHub MCP access without putting a PAT in the user.message.
+2. `POST /v1/sessions/$id/events` with `{events: [{type: "user.message", content: [{type: "text", text: "$MSG"}]}]}` — triggers the agent to start running.
+3. Poll `GET /v1/sessions/$id` every 30s. Return when status is `idle`, `completed`, or `stopped`. Fail on `failed`/`errored`. Time out at 20 min.
+
+Per-session state lands at `tmp/sessions/<label>.id` (session id) and `tmp/logs/<label>.log` (full curl + status trace). For retry, rerun a single label:
+
+```bash
+bash skills/webster-weekly-council/scripts/run-agent-session.sh seo "$SEO_ID" "$MSG_CRITIC"
+```
+
+## Re-run discipline
+
+If one critic fails or times out:
+
+1. Look at `tmp/logs/<label>.log` to see the last status. `failed` ≠ no findings — the critic may have committed before the API errored.
+2. `git fetch origin "$BRANCH" && git diff main --stat` to see what landed.
+3. If findings file is stub or missing, re-run the single label.
+4. Then proceed to phase 5 (verify).
+
+Do NOT re-run the entire fan-out for a single failure. The other 5 critics have already committed; another full run would create duplicate findings.
diff --git a/skills/webster-weekly-council/references/genealogy.md b/skills/webster-weekly-council/references/genealogy.md
new file mode 100644
index 0000000..081121e
--- /dev/null
+++ b/skills/webster-weekly-council/references/genealogy.md
@@ -0,0 +1,76 @@
+# Phase 5.5 — Critic Genealogy (gap-detection, fail-open)
+
+**0–10 min. Fail-open by design.** If genealogy errors, log and continue. Redesigner still runs against the 5 original critic findings.
+
+This is the hero capability. `scripts/critic-genealogy.ts` reviews the five critic outputs on `$BRANCH`. If Opus 4.7 detects a recurring pattern unowned by the existing council, it:
+
+1. Clones the `brand-voice-critic.json` template
+2. Swaps in the new scope
+3. Registers the spec via `POST /v1/agents`
+4. Creates a session with the vault
+5. Invokes the new critic on `$BRANCH`
+6. Commits its findings + spec + session log to `history/$WEEK_DATE/genealogy/`
+
+The redesigner in phase 6 then reads all findings on `$BRANCH` (including the spawned critic's) without further wiring.
+
+## Run
+
+```bash
+echo "=== Step 5.5: critic genealogy ==="
+GENEALOGY_LOG="tmp/logs/genealogy.log"
+: > "$GENEALOGY_LOG"
+
+set +e
+GENEALOGY_ARGS=(
+  --branch "$BRANCH"
+  --week "$WEEK_DATE"
+  --lp-target "$LP_TARGET"
+)
+if [[ -s "$PLANNER_REQUEST_FILE" ]]; then
+  GENEALOGY_ARGS+=(--planner-request "$PLANNER_REQUEST_FILE")
+fi
+
+bun scripts/critic-genealogy.ts "${GENEALOGY_ARGS[@]}" \
+  2>&1 | tee "$GENEALOGY_LOG"
+GENEALOGY_CODE=${PIPESTATUS[0]}
+set -e
+
+if (( GENEALOGY_CODE == 0 )); then
+  if grep -q "no gap detected" "$GENEALOGY_LOG"; then
+    GENEALOGY_STATUS="no gap — council coverage complete"
+    echo "✓ $GENEALOGY_STATUS"
+  elif grep -q "genealogy complete" "$GENEALOGY_LOG"; then
+    SPAWNED=$(grep -oE "gap found: [a-z0-9-]+" "$GENEALOGY_LOG" | head -1 | awk '{print $3}')
+    GENEALOGY_STATUS="spawned ${SPAWNED:-new critic} — findings on branch, feeding redesigner"
+    echo "✓ $GENEALOGY_STATUS"
+  else
+    GENEALOGY_STATUS="returned 0 but output unclear — see tmp/logs/genealogy.log"
+    echo "⚠ $GENEALOGY_STATUS"
+  fi
+else
+  GENEALOGY_STATUS="failed (exit $GENEALOGY_CODE, non-blocking) — see tmp/logs/genealogy.log"
+  echo "⚠ $GENEALOGY_STATUS"
+fi
+
+# If a critic was spawned, it committed findings to $BRANCH via GitHub MCP.
+# Refresh our working tree so subsequent local echoes reflect the freshest state.
+# (The redesigner itself re-fetches via MCP; this reset is for our own visibility.)
+git fetch origin "$BRANCH"
+git reset --hard "origin/$BRANCH"
+```
+
+## Why fail-open
+
+- Genealogy is an emergent-capability demo, not a critical-path step. The council's coverage is "complete enough" with the 5 pre-registered critics.
+- Failure modes (API hiccup, gap-detection LLM hallucination, registration 5xx) shouldn't sink the whole weekly run.
+- `$GENEALOGY_STATUS` carries forward into phase 7's PR body and phase 8's checkpoint, so the failure is captured durably even though the run continues.
+
+## Manual retry
+
+To rerun standalone after the session:
+
+```bash
+bun scripts/critic-genealogy.ts --branch "$BRANCH" --week "$WEEK_DATE" --lp-target "$LP_TARGET"
+```
+
+If a planner-suggested gap exists, add `--planner-request "$PLANNER_REQUEST_FILE"`.
diff --git a/skills/webster-weekly-council/references/planner.md b/skills/webster-weekly-council/references/planner.md
new file mode 100644
index 0000000..639cd5c
--- /dev/null
+++ b/skills/webster-weekly-council/references/planner.md
@@ -0,0 +1,108 @@
+# Phase 3 — Run planner (Opus 4.7)
+
+**3–5 min. Fail-closed.** If this phase errors, halt. Do not fan out critics without `history/$WEEK_DATE/plan.md`.
+
+The planner marshals `history/memory.jsonl`, the two newest `history/<week>/verdict.json` files, and the monitor anomaly report into one planner context. Then invokes `webster-planner` (Opus 4.7), writes `history/$WEEK_DATE/plan.md`, and appends the planner event row to `history/memory.jsonl`.
+
+```bash
+echo "=== Step 3: planner ==="
+PLANNER_LOG="tmp/logs/planner.log"
+PLANNER_CONTEXT_FILE="tmp/planner-context-$WEEK_DATE.txt"
+PLAN_PATH="history/$WEEK_DATE/plan.md"
+: > "$PLANNER_LOG"
+
+if ! bun --eval '
+import { marshalPlannerContext } from "./scripts/planner-context.ts";
+
+const [memoryPath, verdictDir, monitorPath, outputPath] = process.argv.slice(1);
+const contextText = marshalPlannerContext({ memoryPath, verdictDir, monitorPath });
+await Bun.write(outputPath, contextText);
+' \
+  "history/memory.jsonl" \
+  "history" \
+  "context/monitor/alerts.md" \
+  "$PLANNER_CONTEXT_FILE" \
+  >>"$PLANNER_LOG" 2>&1; then
+  echo "ABORT: planner context marshaling failed."
+  echo "See error details in $PLANNER_LOG:"
+  tail -80 "$PLANNER_LOG"
+  exit 1
+fi
+
+set +e
+bun --eval '
+import { readFileSync } from "node:fs";
+import { invokePlanner } from "./scripts/planner-invoke.ts";
+
+const [contextPath, week, historyDir, apiKey] = process.argv.slice(1);
+const { planPath } = await invokePlanner({
+  contextText: readFileSync(contextPath, "utf8"),
+  week,
+  historyDir,
+  apiKey,
+});
+console.log(`planner wrote ${planPath}`);
+' \
+  "$PLANNER_CONTEXT_FILE" \
+  "$WEEK_DATE" \
+  "history" \
+  "$ANTHROPIC_API_KEY" \
+  >>"$PLANNER_LOG" 2>&1
+PLANNER_CODE=$?
+set -e
+
+if (( PLANNER_CODE != 0 )); then
+  echo "ABORT: planner failed before critic fan-out (exit $PLANNER_CODE)."
+  echo "See error details in $PLANNER_LOG:"
+  tail -80 "$PLANNER_LOG"
+  exit "$PLANNER_CODE"
+fi
+
+if [[ ! -s "$PLAN_PATH" ]]; then
+  echo "ABORT: planner completed but $PLAN_PATH is missing or empty."
+  echo "See $PLANNER_LOG for details."
+  exit 1
+fi
+
+git add "$PLAN_PATH" history/memory.jsonl
+git commit -m "chore(planner): add plan for week $WEEK_DATE"
+git push origin "$BRANCH"
+
+PLAN_TEXT=$(cat "$PLAN_PATH")
+PLAN_JSON_FILE="tmp/planner-plan-$WEEK_DATE.json"
+PLANNER_REQUEST_FILE="tmp/planner-new-critic-request-$WEEK_DATE.json"
+
+python3 skills/webster-weekly-council/scripts/extract-planner-json.py \
+  "$PLAN_PATH" "$PLAN_JSON_FILE"
+
+if jq -e 'has("new_critic_request") and (.new_critic_request != null)' "$PLAN_JSON_FILE" >/dev/null; then
+  jq '.new_critic_request' "$PLAN_JSON_FILE" > "$PLANNER_REQUEST_FILE"
+  echo "Planner requested possible new critic: $PLANNER_REQUEST_FILE"
+else
+  rm -f "$PLANNER_REQUEST_FILE"
+fi
+
+echo "✓ Planner output ready for critics: $PLAN_PATH"
+```
+
+## What `plan.md` contains
+
+The planner emits a JSON block (extracted by `extract-planner-json.py`) plus rendered markdown. Schema:
+
+- `classification` — one-line summary of the week's state
+- `next_action` — one of `promote_and_experiment | hold_baseline | revert_and_retry | explore_broadly`
+- `direction_hint` — the weighting input critics see in their fan-out user.message
+- `new_critic_request` — optional; if present, phase 5.5 (genealogy) reads it as a planner-suggested gap
+- `rationale` — natural-language reasoning trace
+
+## Why this phase is fail-closed
+
+- Critics' system prompts read `plan.md` from the branch. No plan = no direction hint = generic critic findings.
+- Memory append happens during `invokePlanner`. If we fan out without it, the next week's planner sees a gap in the event log.
+
+## Outputs ready for phase 4
+
+- `history/$WEEK_DATE/plan.md` (committed + pushed to `$BRANCH`)
+- `$PLAN_TEXT` (in-memory, passed to fan-out user.messages)
+- `$PLANNER_REQUEST_FILE` (optional, consumed by phase 5.5)
+- `history/memory.jsonl` (appended `verdict-ready` row)
diff --git a/skills/webster-weekly-council/references/preflight.md b/skills/webster-weekly-council/references/preflight.md
new file mode 100644
index 0000000..12446a6
--- /dev/null
+++ b/skills/webster-weekly-council/references/preflight.md
@@ -0,0 +1,98 @@
+# Phase 0 — Pre-flight + session constants
+
+Mandatory. Do not skip. Both blocks below must run successfully before any other phase.
+
+## Pre-flight (fail-closed)
+
+```bash
+# 1. ANTHROPIC_API_KEY must NOT be in shell env — it would silently bill
+#    `claude -p` and any Forge call against API credits instead of the Max sub.
+if [[ -n "$ANTHROPIC_API_KEY" ]]; then
+  echo "ABORT: ANTHROPIC_API_KEY is exported in your shell."
+  echo "Fix: exit wbs, run 'unset ANTHROPIC_API_KEY', relaunch wbs, retry this prompt."
+  exit 1
+fi
+
+# 2. Fetch the key from macOS keychain for the duration of this bash scope only.
+ANTHROPIC_API_KEY=$(security find-generic-password -s anthropic-webster -a "$USER" -w 2>/dev/null)
+if [[ -z "$ANTHROPIC_API_KEY" ]]; then
+  echo "ABORT: ANTHROPIC_API_KEY missing from keychain."
+  echo "One-time setup: security add-generic-password -U -s anthropic-webster -a \"\$USER\" -w \"sk-ant-...\""
+  exit 1
+fi
+export ANTHROPIC_API_KEY  # scoped to this bash — the outer wbs process stays clean
+
+# 3. GITHUB_TOKEN is NO LONGER needed in user.message or as env.
+#    The vault holds the PAT; sessions inject it at MCP tool-call time.
+#    Do not check for it, do not use it.
+
+# 4. Registration artifacts must exist (session 1 output).
+REQUIRED=(
+  "environments/webster-council-env.id"
+  "context/monitor/id.txt"
+  "context/critics/brand-voice/id.txt"
+  "context/critics/fh-compliance/id.txt"
+  "context/critics/seo/id.txt"
+  "context/critics/conversion/id.txt"
+  "context/critics/copy/id.txt"
+  "context/redesigner/id.txt"
+)
+MISSING=()
+for f in "${REQUIRED[@]}"; do
+  [[ -s "$f" ]] || MISSING+=("$f")
+done
+if (( ${#MISSING[@]} > 0 )); then
+  echo "ABORT: registration artifacts missing:"
+  printf '  - %s\n' "${MISSING[@]}"
+  echo "Run session 1 first (prompts/first-wbs-session.md), or restore from a prior checkpoint."
+  exit 1
+fi
+
+mkdir -p tmp/logs tmp/sessions
+```
+
+## Session constants
+
+```bash
+WEEK_DATE=$(date -u +%Y-%m-%d)
+relative_utc_date() {
+  local offset="$1"
+  local format="${2:-+%Y-%m-%d}"
+  if date -u -d "$offset" "$format" >/dev/null 2>&1; then
+    date -u -d "$offset" "$format"
+  else
+    date -u -v"$offset" "$format"
+  fi
+}
+PREV_WEEK_DATE=$(relative_utc_date "-7d")
+BRANCH="council/$WEEK_DATE"
+LP_TARGET="https://certified.richerhealth.ca"
+VAULT_ID="vlt_011CaLe2pEofWQptxQyV4UMd"   # webster vault, holds GitHub PAT
+ENV_ID=$(cat environments/webster-council-env.id)
+
+# Agent IDs
+MONITOR_ID=$(cat context/monitor/id.txt)
+BRAND_VOICE_ID=$(cat context/critics/brand-voice/id.txt)
+FH_COMPLIANCE_ID=$(cat context/critics/fh-compliance/id.txt)
+SEO_ID=$(cat context/critics/seo/id.txt)
+CONVERSION_ID=$(cat context/critics/conversion/id.txt)
+COPY_ID=$(cat context/critics/copy/id.txt)
+REDESIGNER_ID=$(cat context/redesigner/id.txt)
+
+API="https://api.anthropic.com/v1"
+BETA_HDR="anthropic-beta: managed-agents-2026-04-01"
+VERSION_HDR="anthropic-version: 2023-06-01"
+
+echo "WEEK_DATE=$WEEK_DATE  BRANCH=$BRANCH  LP_TARGET=$LP_TARGET"
+echo "ENV_ID=$ENV_ID  VAULT_ID=$VAULT_ID"
+```
+
+These variables are referenced by every later phase. Keep this bash scope alive (don't re-launch the shell) until publish completes.
+
+## Why this phase is fail-closed
+
+- Missing API key → silent billing catastrophe.
+- Missing registration artifacts → fan-out creates sessions that error mid-run, wasting tokens before the operator notices.
+- Wrong `WEEK_DATE` → `history/<date>/` collisions or empty diff against prior week.
+
+If any check fails, fix the named prerequisite and re-run preflight. The script is idempotent.
diff --git a/skills/webster-weekly-council/references/publish.md b/skills/webster-weekly-council/references/publish.md
new file mode 100644
index 0000000..8254e22
--- /dev/null
+++ b/skills/webster-weekly-council/references/publish.md
@@ -0,0 +1,109 @@
+# Phase 7 + 8 — Open draft PR + write checkpoint
+
+**~1 min.** Package the council run as a draft PR for human review, then write a durable completion checkpoint.
+
+## Phase 7 — Open draft PR
+
+````bash
+if gh pr view "$BRANCH" --json number 2>/dev/null | jq -e '.number' >/dev/null; then
+  PR_URL=$(gh pr view "$BRANCH" --json url --jq '.url')
+  echo "PR already exists for $BRANCH: $PR_URL"
+else
+  PR_BODY=$(cat <<EOF
+## Council run — week $WEEK_DATE
+
+Automated planner + fan-out: planner wrote this week's direction, then monitor + 5 critics + redesigner ran. Findings committed to this branch, synthesized into a weekly redesign proposal.
+
+## Findings
+
+$(printf -- '- %s\n' "${SUCCEEDED[@]}")
+
+## Critic Genealogy
+
+${GENEALOGY_STATUS:-not run}
+
+## Planner output
+
+- \`$PLAN_PATH\`
+
+## Redesigner output
+
+- \`$PROPOSAL\`
+- \`$DECISION\`
+
+$([[ -f "$DECISION" ]] && echo "## Decision summary" && echo "" && echo '```json' && cat "$DECISION" && echo '```')
+
+---
+
+Opened automatically by the Webster dispatcher. Merge when ready.
+EOF
+)
+  PR_URL=$(gh pr create --draft \
+    --base main --head "$BRANCH" \
+    --title "council: week $WEEK_DATE proposal" \
+    --body "$PR_BODY" --json url --jq '.url' 2>/dev/null \
+    || gh pr create --draft \
+         --base main --head "$BRANCH" \
+         --title "council: week $WEEK_DATE proposal" \
+         --body "$PR_BODY" | tail -1)
+fi
+echo "PR: $PR_URL"
+````
+
+The double `gh pr create` form covers gh CLIs that don't support `--json` on `create` — first attempt with JSON, fall back to plain.
+
+## Phase 8 — Checkpoint + exit
+
+```bash
+CKPT=".claude/checkpoints/$(date -u +%Y-%m-%dT%H%M%SZ)-session-2-complete.md"
+mkdir -p .claude/checkpoints
+cat > "$CKPT" <<EOF
+---
+ts: $(date -u +%Y-%m-%dT%H:%M:%SZ)
+trigger: session-2-complete
+---
+
+## What happened
+
+Full council run for week $WEEK_DATE. Ran planner first, then monitor + 5 critics in parallel, then redesigner. All artifacts committed to \`$BRANCH\`. Draft PR opened.
+
+## Results
+
+- Planner output: $PLAN_PATH
+- Findings with content: ${#SUCCEEDED[@]}/${#FILES[@]}
+- Critic Genealogy: ${GENEALOGY_STATUS:-not run}
+- Redesigner completed: $([[ $REDESIGNER_RESULT -eq 0 ]] && echo yes || echo "no (exit $REDESIGNER_RESULT)")
+- PR: $PR_URL
+- Mock history seeded: $($SEED_NEEDED && echo "yes (10 weeks + current)" || echo "no (already present)")
+
+## Live state
+
+- Branch: $BRANCH
+- plan.md: $([[ -f "$PLAN_PATH" ]] && echo committed || echo MISSING)
+- proposal.md: $([[ -f "$PROPOSAL" ]] && echo committed || echo MISSING)
+- decision.json: $([[ -f "$DECISION" ]] && echo committed || echo MISSING)
+
+## Next tick
+
+$( (( ${#STUBBED[@]} > 0 )) && echo "- Re-run the stubbed critics manually and amend the PR:" && printf '  - %s\n' "${STUBBED[@]}" )
+- Fork certified.richerhealth.ca into \`site/\` so the next council can produce a real diff instead of a brief.
+- Record the demo video showing: fan-out → genealogy → synthesis → PR.
+EOF
+
+git add "$CKPT"
+git commit -m "docs(checkpoint): session 2 complete — council week $WEEK_DATE"
+git push origin "$BRANCH"
+
+echo ""
+echo "══════════════════════════════════════════════════"
+echo "Session 2 complete."
+echo "  PR: $PR_URL"
+echo "  Checkpoint: $CKPT"
+echo "══════════════════════════════════════════════════"
+exit 0
+```
+
+## What this phase explicitly does NOT do
+
+- **Does not merge the PR.** The PR is opened as a draft. Human review = approval.
+- **Does not fork `site/`.** Without `site/`, the redesigner produces a brief (`proposal.md`) instead of a diff (`proposal.diff`). That is by design for this session — `site/` fork belongs in its own focused session (see `context/SITE-FORK-CHECKLIST.md`).
diff --git a/skills/webster-weekly-council/references/redesign.md b/skills/webster-weekly-council/references/redesign.md
new file mode 100644
index 0000000..03d69dd
--- /dev/null
+++ b/skills/webster-weekly-council/references/redesign.md
@@ -0,0 +1,63 @@
+# Phase 6 — Redesigner session
+
+**5–10 min.** Single synthesis call. Reads all findings on `$BRANCH`, commits `history/$WEEK_DATE/proposal.md` + `decision.json` to the same branch.
+
+```bash
+MSG_REDESIGNER="BRANCH=$BRANCH
+WEEK_DATE=$WEEK_DATE
+PLAN_PATH=$PLAN_PATH
+
+Planner context (additive only; use direction_hint as a weighting input, but critic vetoes and CRITICAL/HIGH evidence remain sovereign):
+$PLAN_TEXT"
+
+bash skills/webster-weekly-council/scripts/run-agent-session.sh \
+  redesigner "$REDESIGNER_ID" "$MSG_REDESIGNER"
+REDESIGNER_RESULT=$?
+
+if (( REDESIGNER_RESULT != 0 )); then
+  echo "WARNING: redesigner did not complete cleanly (exit $REDESIGNER_RESULT)."
+  echo "Findings are still committed on $BRANCH — you can run the redesigner manually later."
+fi
+
+git fetch origin "$BRANCH"
+git reset --hard "origin/$BRANCH"
+
+PROPOSAL="history/$WEEK_DATE/proposal.md"
+DECISION="history/$WEEK_DATE/decision.json"
+if [[ -f "$PROPOSAL" && -f "$DECISION" ]]; then
+  echo "✓ Redesigner output:"
+  echo "    $PROPOSAL"
+  echo "    $DECISION"
+else
+  echo "WARNING: redesigner output missing on $BRANCH."
+  echo "  proposal.md: $([[ -f "$PROPOSAL" ]] && echo present || echo MISSING)"
+  echo "  decision.json: $([[ -f "$DECISION" ]] && echo present || echo MISSING)"
+fi
+```
+
+## What the redesigner reads
+
+- `history/$WEEK_DATE/plan.md` — direction hint, classification, next-action
+- `context/critics/*/findings.md` — five pre-registered critics' output
+- `context/monitor/alerts.md` — anomaly report
+- `history/$WEEK_DATE/genealogy/*` if genealogy spawned a critic — its findings + spec
+
+## Sovereignty rule
+
+Planner direction is a weighting input. If a critic flagged a CRITICAL or HIGH issue outside the planner's direction, the redesigner must still address it. Findings vetoes plan; plan never vetoes findings.
+
+## What it writes
+
+- `proposal.md` — the redesign brief (prose + recommendations, structured by severity)
+- `decision.json` — machine-readable summary used by phase 7 (PR body) and phase 8 (checkpoint), and read by next week's planner via `history/memory.jsonl`
+
+## If the redesigner fails
+
+Findings are already on `$BRANCH`. Re-run just this phase:
+
+```bash
+bash skills/webster-weekly-council/scripts/run-agent-session.sh \
+  redesigner "$REDESIGNER_ID" "$MSG_REDESIGNER"
+```
+
+The PR can still be opened in phase 7 with the findings alone; redesigner output becomes "MISSING" in the PR body.
diff --git a/skills/webster-weekly-council/references/seed-history.md b/skills/webster-weekly-council/references/seed-history.md
new file mode 100644
index 0000000..2456829
--- /dev/null
+++ b/skills/webster-weekly-council/references/seed-history.md
@@ -0,0 +1,74 @@
+# Phase 1 — Seed 10-week mock analytics history
+
+**Idempotent. ~2 min. Skipped if all 10 weeks already present.**
+
+Monitor needs prior-week baselines to compute WoW deltas. If history is missing, the monitor writes "Week 1 baseline: no prior week" and skips anomaly detection — survivable but wastes monitor spend.
+
+Plausible small-wellness-biz shape with a slight downtrend in the most recent 2 weeks so the monitor fires on CURRENT vs PREV comparison.
+
+```bash
+SEED_NEEDED=false
+for i in $(seq 1 10); do
+  DATE=$(relative_utc_date "-${i}w")
+  [[ -f "history/$DATE/analytics.json" ]] || { SEED_NEEDED=true; break; }
+done
+
+if $SEED_NEEDED; then
+  echo "Seeding 10 weeks of mock analytics..."
+  for i in $(seq 1 10); do
+    DATE=$(relative_utc_date "-${i}w")
+    mkdir -p "history/$DATE"
+
+    # Recent 2 weeks: dip; older 8: baseline
+    if [[ $i -le 2 ]]; then
+      SESSIONS=$((900 + RANDOM % 200))    # ~900–1100 — recent dip
+      CONV=$((7 + RANDOM % 4))             # 7–10 — recent drop
+      BOUNCE=$((62 + RANDOM % 6))          # 62–68% — recent rise
+    else
+      SESSIONS=$((1100 + RANDOM % 300))   # ~1100–1400 — baseline
+      CONV=$((10 + RANDOM % 6))            # 10–15 — baseline
+      BOUNCE=$((50 + RANDOM % 8))          # 50–58% — baseline
+    fi
+
+    jq -n \
+      --arg date "$DATE" \
+      --arg sessions "$SESSIONS" \
+      --arg conv "$CONV" \
+      --arg bounce "$BOUNCE" \
+      '{
+        week_ending: $date,
+        sessions: ($sessions | tonumber),
+        conversions: ($conv | tonumber),
+        bounce_rate_pct: ($bounce | tonumber),
+        traffic_sources: {
+          organic: 0.50,
+          direct: 0.25,
+          social: 0.15,
+          referral: 0.10
+        }
+      }' > "history/$DATE/analytics.json"
+  done
+
+  # Add THIS week's file too so the monitor has something to compare against
+  mkdir -p "history/$WEEK_DATE"
+  jq -n --arg date "$WEEK_DATE" '{
+    week_ending: $date,
+    sessions: 850,
+    conversions: 6,
+    bounce_rate_pct: 67,
+    traffic_sources: { organic: 0.42, direct: 0.28, social: 0.18, referral: 0.12 }
+  }' > "history/$WEEK_DATE/analytics.json"
+
+  git add history/
+  git commit -m "chore(history): seed 10 weeks mock analytics + week $WEEK_DATE baseline"
+  git push origin main
+else
+  echo "History already present; skipping seed."
+fi
+```
+
+## Notes
+
+- `RANDOM` is bash's built-in PRNG. Reseeding produces different numbers — that's fine; the seeder runs once.
+- The seeder commits to `main`, not `$BRANCH`. History is shared across all weekly runs.
+- If the loop finds even one missing week, it re-seeds all 10 + this week. That keeps the seed coherent (same shape across all 10 weeks).
diff --git a/skills/webster-weekly-council/references/verify.md b/skills/webster-weekly-council/references/verify.md
new file mode 100644
index 0000000..a04c0fc
--- /dev/null
+++ b/skills/webster-weekly-council/references/verify.md
@@ -0,0 +1,63 @@
+# Phase 5 — Verify findings (redesigner gate)
+
+**~2 min.** Confirm enough real findings landed before spending redesigner tokens.
+
+```bash
+git fetch origin "$BRANCH"
+git reset --hard "origin/$BRANCH"
+
+FILES=(
+  context/monitor/alerts.md
+  context/critics/brand-voice/findings.md
+  context/critics/fh-compliance/findings.md
+  context/critics/seo/findings.md
+  context/critics/conversion/findings.md
+  context/critics/copy/findings.md
+)
+
+SUCCEEDED=()
+STUBBED=()
+for f in "${FILES[@]}"; do
+  if grep -q "No runs yet\." "$f" 2>/dev/null; then
+    STUBBED+=("$f")
+  else
+    SUCCEEDED+=("$f")
+  fi
+done
+
+echo "Findings with content: ${#SUCCEEDED[@]}/${#FILES[@]}"
+printf '  ✓ %s\n' "${SUCCEEDED[@]}"
+if (( ${#STUBBED[@]} > 0 )); then
+  printf '  ✗ %s (still stub)\n' "${STUBBED[@]}"
+fi
+
+# Abort redesigner if too few critics contributed
+if (( ${#SUCCEEDED[@]} < 3 )); then
+  echo "ABORT: only ${#SUCCEEDED[@]} findings files have content. Need ≥3 for redesigner."
+  echo "Check tmp/logs/ for what went wrong; fix and re-run from phase 4."
+  exit 1
+fi
+
+# Run our own gate against what the critics wrote — catches format drift
+bun scripts/validate-findings.ts
+```
+
+## Why the gate is `>=3 critics`
+
+- Redesigner synthesizes findings into a single proposal. Below 3 contributors, the proposal collapses to one critic's voice with no cross-cutting signal.
+- Stub detection looks for the literal phrase `No runs yet.` — that's the placeholder string critics' system prompts overwrite when they run. Any other content is treated as a real finding.
+
+## What `validate-findings.ts` checks
+
+- File exists and is non-empty
+- `# Findings — Week YYYY-MM-DD` header present
+- At least one `[CRITICAL|HIGH|MEDIUM|LOW]` severity tag
+- Out-of-scope section present (even if empty)
+
+Format drift here is a data-quality bug — don't paper over a failing validate by editing critic output. Re-run the offending critic.
+
+## Outputs ready for phase 5.5
+
+- `$SUCCEEDED` array — used in phase 7 PR body
+- `$STUBBED` array — used in phase 8 checkpoint "next tick" section
+- Working tree synced to `origin/$BRANCH` (so subsequent local echoes reflect freshest state)
diff --git a/skills/webster-weekly-council/scripts/extract-planner-json.py b/skills/webster-weekly-council/scripts/extract-planner-json.py
new file mode 100755
index 0000000..dffcd39
--- /dev/null
+++ b/skills/webster-weekly-council/scripts/extract-planner-json.py
@@ -0,0 +1,26 @@
+#!/usr/bin/env python3
+"""extract-planner-json.py <plan.md> <plan.json>
+
+Pull the JSON block out of `history/<week>/plan.md` so phase 3 can run
+`jq -e 'has("new_critic_request")' plan.json` without parsing markdown in
+bash. The planner emits both rendered markdown and a fenced ```json block;
+this extractor returns the JSON. If no fenced block is found the whole file
+is treated as JSON (matches behavior of the original inline heredoc in
+`prompts/second-wbs-session.md`).
+"""
+
+import re
+import sys
+
+if len(sys.argv) != 3:
+    print(f"usage: {sys.argv[0]} <plan.md> <plan.json>", file=sys.stderr)
+    sys.exit(1)
+
+with open(sys.argv[1], encoding="utf-8") as src:
+    plan_text = src.read()
+
+match = re.search(r"```(?:json)?\s*(\{.*?\})\s*```", plan_text, re.S)
+json_text = match.group(1) if match else plan_text
+
+with open(sys.argv[2], "w", encoding="utf-8") as dst:
+    _ = dst.write(json_text)
diff --git a/skills/webster-weekly-council/scripts/run-agent-session.sh b/skills/webster-weekly-council/scripts/run-agent-session.sh
new file mode 100755
index 0000000..79c44bf
--- /dev/null
+++ b/skills/webster-weekly-council/scripts/run-agent-session.sh
@@ -0,0 +1,85 @@
+#!/usr/bin/env bash
+# run-agent-session.sh <label> <agent_id> <user_message>
+#
+# Create a Webster managed-agent session attached to the council vault, send
+# one user.message event to start it, poll until idle/completed/timeout.
+#
+# Exit codes:
+#   0 — agent reached idle/completed/stopped
+#   1 — failure (could not create session, agent reported failed/errored, or
+#       missing required env)
+#   2 — timeout (20-min wall-clock without idling)
+#
+# Required env (caller exports these from references/preflight.md):
+#   ANTHROPIC_API_KEY  — fetched from macOS keychain by preflight, scoped to
+#                        the caller's bash; not the operator shell env
+#   ENV_ID             — environments/webster-council-env.id contents
+#   VAULT_ID           — vlt_011CaLe2pEofWQptxQyV4UMd (webster vault)
+#   WEEK_DATE          — UTC date used in session title
+#   API                — https://api.anthropic.com/v1
+#   VERSION_HDR        — "anthropic-version: 2023-06-01"
+#   BETA_HDR           — "anthropic-beta: managed-agents-2026-04-01"
+#
+# Outputs:
+#   tmp/sessions/<label>.id  — session id (single line, no trailing newline)
+#   tmp/logs/<label>.log     — full curl + status trace
+
+set -uo pipefail
+
+LABEL="${1:?label required (e.g. monitor, brand-voice, redesigner)}"
+AGENT_ID="${2:?agent_id required}"
+MSG="${3:?user_message required}"
+
+for var in ANTHROPIC_API_KEY ENV_ID VAULT_ID WEEK_DATE API VERSION_HDR BETA_HDR; do
+  if [[ -z "${!var:-}" ]]; then
+    echo "ERROR: missing required env: $var" >&2
+    exit 1
+  fi
+done
+
+LOG="tmp/logs/$LABEL.log"
+IDFILE="tmp/sessions/$LABEL.id"
+mkdir -p tmp/logs tmp/sessions
+: > "$LOG"
+
+# 1. Create session attached to vault
+SESSION_ID=$(curl -fsS "$API/sessions" \
+  -H "x-api-key: $ANTHROPIC_API_KEY" \
+  -H "$VERSION_HDR" -H "$BETA_HDR" \
+  -H "content-type: application/json" \
+  -d "$(jq -n --arg a "$AGENT_ID" --arg e "$ENV_ID" --arg v "$VAULT_ID" --arg t "$LABEL $WEEK_DATE" \
+        '{agent: $a, environment_id: $e, vault_ids: [$v], title: $t}')" \
+  2>>"$LOG" | jq -r '.id')
+
+if [[ -z "$SESSION_ID" || "$SESSION_ID" == "null" ]]; then
+  echo "ERROR: failed to create session for $LABEL" | tee -a "$LOG" >&2
+  exit 1
+fi
+printf '%s' "$SESSION_ID" > "$IDFILE"
+echo "[$LABEL] session=$SESSION_ID" >> "$LOG"
+
+# 2. Send user.message (triggers agent to start running)
+curl -fsS "$API/sessions/$SESSION_ID/events" \
+  -H "x-api-key: $ANTHROPIC_API_KEY" \
+  -H "$VERSION_HDR" -H "$BETA_HDR" \
+  -H "content-type: application/json" \
+  -d "$(jq -n --arg text "$MSG" '{events: [{type: "user.message", content: [{type: "text", text: $text}]}]}')" \
+  >>"$LOG" 2>&1
+
+# 3. Poll every 30s until idle OR 20-min timeout
+DEADLINE=$(( $(date +%s) + 1200 ))
+while (( $(date +%s) < DEADLINE )); do
+  STATUS=$(curl -fsS "$API/sessions/$SESSION_ID" \
+    -H "x-api-key: $ANTHROPIC_API_KEY" \
+    -H "$VERSION_HDR" -H "$BETA_HDR" \
+    2>>"$LOG" | jq -r '.status // "unknown"')
+  echo "[$LABEL] status=$STATUS $(date -u +%T)" >> "$LOG"
+  case "$STATUS" in
+    idle|completed|stopped) exit 0 ;;
+    failed|errored)         echo "[$LABEL] agent failed" >> "$LOG"; exit 1 ;;
+  esac
+  sleep 30
+done
+
+echo "[$LABEL] TIMEOUT after 20min" >> "$LOG"
+exit 2

From 5f450547da2c7382b68975106c5029ba73ee7803 Mon Sep 17 00:00:00 2001
From: richsak <richie@richiesakhon.com>
Date: Sun, 26 Apr 2026 15:55:06 -0700
Subject: [PATCH 2/2] docs: route operators to /webster-weekly-council

Update README, AGENTS, context/ARCHITECTURE, context/FEATURES, and
history/AGENTS so the weekly-run operator surface is the new library
skill (slash-command form) with prompts/second-wbs-session.md framed as
the locked single-page runbook fallback. The prompt itself is unchanged
and still locked by scripts/__tests__/sim-council.test.ts.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
---
 AGENTS.md               | 11 +++++-----
 README.md               | 19 ++++++++++++-----
 context/ARCHITECTURE.md | 21 ++++++++++---------
 context/FEATURES.md     | 46 ++++++++++++++++++++---------------------
 4 files changed, 54 insertions(+), 43 deletions(-)

diff --git a/AGENTS.md b/AGENTS.md
index 797fc51..fc40374 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -13,7 +13,7 @@ This file is for implementation operators. See `skills/webster-lp-audit/SKILL.md
 
 Two active workstreams:
 
-- **Production Webster** — Nicolette's weekly landing-page improvement council runs on `main` via `prompts/second-wbs-session.md`. This is live for her business; do not break it.
+- **Production Webster** — Nicolette's weekly landing-page improvement council runs on `main`. Operator surface: `/webster-weekly-council` (skill at `skills/webster-weekly-council/SKILL.md`) or the single-page runbook at `prompts/second-wbs-session.md`. Both produce identical artifacts; the prompt is the locked source-of-truth runbook. This is live for her business; do not break it.
 - **Hackathon expansion** — Single-substrate Richer Health LP demo with a simulation runner producing timelapse assets. Deadline **2026-04-28**. Working branch: `dev/`. See `context/VISION.md` for canonical north-star.
 
 ## First actions every session
@@ -126,12 +126,13 @@ Use `TaskCreate` / `TaskUpdate` for multi-step work within a single session. Tas
 
 ## Skill invocation (Claude Code)
 
-Webster ships two runtime-critic skills:
+Webster ships these skills:
 
-- `skills/webster-lp-audit/SKILL.md` — shared council run flow (referenced by production critics)
-- `skills/webster-onboarding/SKILL.md` — end-user onboarding flow (universal, demo placeholder)
+- `skills/webster-weekly-council/SKILL.md` — operator surface for the weekly run. Library skill: SKILL.md index + on-demand phase references + helper scripts. Slash-command form: `/webster-weekly-council`. Equivalent single-page runbook at `prompts/second-wbs-session.md`.
+- `skills/webster-lp-audit/SKILL.md` — shared council run discipline (referenced by production critics)
+- `skills/webster-browser-audit/SKILL.md` — headless browser audit for visual review
 
-If your work modifies either skill, test with a sample invocation before committing.
+If your work modifies any skill, test with a sample invocation before committing. The weekly-council skill must stay artifact-equivalent with `prompts/second-wbs-session.md` — when in doubt, fix the skill, never the prompt.
 
 ## Parallel stream etiquette
 
diff --git a/README.md b/README.md
index d38dced..72a719e 100644
--- a/README.md
+++ b/README.md
@@ -79,7 +79,7 @@ bun scripts/critic-genealogy.ts --fixtures scripts/__tests__/fixtures/genealogy
 
 **30-second pitch:** Webster is an autonomous landing-page improvement council. Nine Claude Managed Agents plan, audit, monitor, synthesize, and package one weekly redesign proposal; the standout demo is Critic Genealogy, where Opus 4.7 detects an unowned audit gap and registers a new specialist at runtime.
 
-**Live-run evidence:** the full operator path is [`prompts/second-wbs-session.md`](prompts/second-wbs-session.md), registration IDs live in `environments/webster-council-env.id` and `context/*/id.txt`, and run artifacts are written under `history/<week>/` when the weekly prompt is executed.
+**Live-run evidence:** the operator surface is the [`/webster-weekly-council`](skills/webster-weekly-council/SKILL.md) skill (library: SKILL.md index + on-demand phase references + helper scripts); the full single-page runbook lives at [`prompts/second-wbs-session.md`](prompts/second-wbs-session.md). Registration IDs live in `environments/webster-council-env.id` and `context/*/id.txt`. Run artifacts are written under `history/<week>/` when the weekly run executes.
 
 **Demo arc artifacts:** the hackathon timelapse animates an 11-week simulation council run. Per-week deliverables live under [`demo-output/landing-page/`](demo-output/landing-page/) (`w00..w10`): desktop/mobile/tablet screenshots, heatmap JSON+SVG, synthetic analytics, and the visual reviewer's markdown verdict. Anthropic Managed Agents memory-store provisioning is captured at [`assets/memory-stores-screenshots/`](assets/memory-stores-screenshots/). The rendered timelapse is hosted externally (link in the submission form); reproduce locally with `bun skills/webster-video/scripts/hydrate-demo-assets.ts && cd video && npx hyperframes render -q high --strict`.
 
@@ -96,9 +96,10 @@ webster/
 │   └── simulation/      9 LP-sim specs (1:1 mirror) that drive the timelapse demo
 ├── context/             architecture, features, quality gates, per-critic findings dirs
 ├── environments/        webster-council-env.json (single Anthropic environment)
-├── prompts/             first-wbs-session.md (bootstrap), second-wbs-session.md (weekly run)
+├── prompts/             first-wbs-session.md (bootstrap), second-wbs-session.md (weekly run runbook)
 ├── scripts/             validate-agents, validate-findings, critic-genealogy
-├── skills/              webster-lp-audit (shared critic discipline), webster-onboarding
+├── skills/              webster-weekly-council (operator surface for the weekly run),
+│                        webster-lp-audit (shared critic discipline)
 ├── .github/workflows/   CI: type + lint + format + schema + findings + markdown + tests
 ├── .husky/              pre-commit runs the same gates locally
 └── AGENTS.md            operator guide for in-repo work
@@ -106,7 +107,7 @@ webster/
 
 ## The weekly flow
 
-The live council runner is a bash-in-markdown prompt: [`prompts/second-wbs-session.md`](prompts/second-wbs-session.md). It:
+The live council runner is a Claude Code library skill: [`/webster-weekly-council`](skills/webster-weekly-council/SKILL.md) — slim SKILL.md index, on-demand phase references under `references/`, and reusable helper scripts under `scripts/`. The single-page bash-in-markdown runbook at [`prompts/second-wbs-session.md`](prompts/second-wbs-session.md) is the same flow as a scrollable readable page. Both produce identical artifacts. The flow:
 
 1. Seeds 10 weeks of mock analytics on first run (monitor needs baselines to diff).
 2. Prepares a shared `council/YYYY-MM-DD` branch.
@@ -169,11 +170,19 @@ Registers the single environment + 9 production agents against the Anthropic API
 
 ### Weekly council run
 
+In Claude Code (primary):
+
+```text
+/webster-weekly-council
+```
+
+Or as a single-page prompt (fallback):
+
 ```bash
 wbs @prompts/second-wbs-session.md
 ```
 
-Runs the full planner + fan-out + redesigner + draft PR described above.
+Both run the full planner + fan-out + redesigner + draft PR described above. The skill loads phase references on demand (smaller per-turn context budget); the prompt is one readable file.
 
 ### Spawn a genealogy critic manually
 
diff --git a/context/ARCHITECTURE.md b/context/ARCHITECTURE.md
index 2d9be14..0493ab8 100644
--- a/context/ARCHITECTURE.md
+++ b/context/ARCHITECTURE.md
@@ -2,7 +2,7 @@
 
 > Mirrors [[webster-architecture]] in vault. Canonical source is this file for in-repo operators; vault file for cross-session memory.
 >
-> **Submission state**: Layers 1–4 + Layer 7 shipped. Layer 5 (`site/` fork + analytics pixel + `scripts/seed-mock-history.ts`) is scoped out for submission — the mock seeder is inlined in `prompts/second-wbs-session.md` Step 1 instead of a separate script, and the redesigner emits `proposal.md` instead of `proposal.diff`. Layer 6 (video) is blocked on Richie's voice record. See `context/FEATURES.md` for per-row status.
+> **Submission state**: Layers 1–4 + Layer 7 shipped. Layer 5 (`site/` fork + analytics pixel + `scripts/seed-mock-history.ts`) is scoped out for submission — the mock seeder is in phase 1 of the weekly-council skill (`skills/webster-weekly-council/references/seed-history.md`) and equivalently in `prompts/second-wbs-session.md` Step 1, and the redesigner emits `proposal.md` instead of `proposal.diff`. Layer 6 (video) is blocked on Richie's voice record. See `context/FEATURES.md` for per-row status.
 
 ## System Overview
 
@@ -48,8 +48,9 @@
 
 ### Layer 1: Routine + Orchestrator
 
-- `routines/weekly-lp-improve.yaml` — cut from submission; weekly trigger is manual `wbs @prompts/second-wbs-session.md`
-- `prompts/second-wbs-session.md` — bash-in-markdown orchestrator (replaces the planned `webster/orchestrator.ts`), reads state, fans out, runs genealogy, opens PR
+- `routines/weekly-lp-improve.yaml` — cut from submission; weekly trigger is manual `/webster-weekly-council` (Claude Code skill) or `wbs @prompts/second-wbs-session.md` (single-page runbook)
+- `skills/webster-weekly-council/SKILL.md` — library skill: slim index + on-demand phase references + helper scripts; the operator surface for the weekly run
+- `prompts/second-wbs-session.md` — bash-in-markdown orchestrator (replaces the planned `webster/orchestrator.ts`), reads state, fans out, runs genealogy, opens PR; the immutable single-page runbook the skill mirrors
 - Shared agent skill `skills/webster-lp-audit/SKILL.md` — universal e2e flow: _read context → critique → write findings → exit_
 - Per-critic context: `context/critics/{name}/findings.md`
 - Run artifacts: `history/YYYY-MM-DD/{analytics.json, council-output/, synthesis.md, proposal.md, decision.json}`
@@ -67,13 +68,13 @@ Environment `environments/webster-council-env.json`:
 
 Agent specs (JSON, not YAML — matches `POST /v1/agents` schema):
 
-- `agents/webster-monitor.json` — Haiku 4.5
-- `agents/brand-voice-critic.json` — Sonnet 4.6
-- `agents/fh-compliance-critic.json` — Sonnet 4.6
-- `agents/seo-critic.json` — Sonnet 4.6
-- `agents/conversion-critic.json` — Sonnet 4.6
-- `agents/copy-critic.json` — Sonnet 4.6
-- `agents/webster-redesigner.json` — Opus 4.7
+- `agents/production/webster-monitor.json` — Haiku 4.5
+- `agents/production/brand-voice-critic.json` — Sonnet 4.6
+- `agents/production/fh-compliance-critic.json` — Sonnet 4.6
+- `agents/production/seo-critic.json` — Sonnet 4.6
+- `agents/production/conversion-critic.json` — Sonnet 4.6
+- `agents/production/copy-critic.json` — Sonnet 4.6
+- `agents/production/webster-redesigner.json` — Opus 4.7
 
 Each spec has: `name`, `model`, `system` (multi-line string with escaped \n), `tools: [{type: agent_toolset_20260401}]`, `metadata`. **No `callable_agents`** (research preview).
 
diff --git a/context/FEATURES.md b/context/FEATURES.md
index 104feb5..be12628 100644
--- a/context/FEATURES.md
+++ b/context/FEATURES.md
@@ -18,7 +18,7 @@
 - **Cut**: 7 (out of submission scope; rationale inline)
 - **Todo**: 7 (1 submission form; all remaining implementation rows shipped or non-implementation blocked/cut)
 
-Hero feature (Critic Genealogy) shipped with live Opus 4.7 validation. All 7 Managed Agents registered. Council fan-out + redesigner + PR automation scripted in `prompts/second-wbs-session.md`. CI green, 29 tests pass. Two scope reassignments below (critic-flow skill renamed; orchestrator moved from TS to bash-in-markdown prompt) — both ship equivalent functionality.
+Hero feature (Critic Genealogy) shipped with live Opus 4.7 validation. All 7 Managed Agents registered. Council fan-out + redesigner + PR automation scripted in `prompts/second-wbs-session.md` (single-page runbook) and exposed as `/webster-weekly-council` library skill at `skills/webster-weekly-council/SKILL.md`. CI green, 29 tests pass. Two scope reassignments below (critic-flow skill renamed; orchestrator moved from TS to bash-in-markdown prompt + library skill) — both ship equivalent functionality.
 
 ## Stream allocation
 
@@ -28,22 +28,22 @@ See `AGENTS.md` for stream → operator mapping.
 
 ## Layer 1: Routine + Orchestrator (Stream 1 — Claude Code Opus 4.7)
 
-| #   | Status | Feature                                                                                                                                                         | Hours |
-| --- | ------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----- |
-| 1   | cut    | `routines/weekly-lp-improve.yaml` — Claude Code Routine with weekly cron. Submission uses manual `wbs @prompts/...`                                             | 2     |
-| 2   | done   | Orchestrator — shipped as `prompts/second-wbs-session.md` (bash-in-markdown, not `.ts`). Functionally equivalent                                                | 4     |
-| 3   | done   | Shared critic skill — shipped as `skills/webster-lp-audit/SKILL.md` (renamed from `critic-flow`)                                                                | 2     |
-| 4   | done   | Per-critic context pattern: `context/critics/{name}/findings.md` (5 critics + monitor seeded)                                                                   | 1     |
-| 5   | done   | Run-artifact pattern: `history/YYYY-MM-DD/` — live `history/2026-04-23/` artifacts include analytics, proposal, decision, operator decision, and genealogy logs | 2     |
-| 6   | done   | Branch + PR automation via `gh pr create` — wired in Step 6 of `second-wbs-session.md`                                                                          | 2     |
+| #   | Status | Feature                                                                                                                                                                                                             | Hours |
+| --- | ------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----- |
+| 1   | cut    | `routines/weekly-lp-improve.yaml` — Claude Code Routine with weekly cron. Submission uses manual `wbs @prompts/...`                                                                                                 | 2     |
+| 2   | done   | Orchestrator — shipped as `prompts/second-wbs-session.md` (bash-in-markdown, not `.ts`) + `/webster-weekly-council` library skill (SKILL.md index + 9 phase references + 2 helper scripts). Functionally equivalent | 4     |
+| 3   | done   | Shared critic skill — shipped as `skills/webster-lp-audit/SKILL.md` (renamed from `critic-flow`)                                                                                                                    | 2     |
+| 4   | done   | Per-critic context pattern: `context/critics/{name}/findings.md` (5 critics + monitor seeded)                                                                                                                       | 1     |
+| 5   | done   | Run-artifact pattern: `history/YYYY-MM-DD/` — live `history/2026-04-23/` artifacts include analytics, proposal, decision, operator decision, and genealogy logs                                                     | 2     |
+| 6   | done   | Branch + PR automation via `gh pr create` — wired in Step 6 of `second-wbs-session.md`                                                                                                                              | 2     |
 
 ## Layer 2: Managed Agent Critics (Stream 2 — Codex heartbeat)
 
 | #   | Status | Feature                                                                                                              | Hours |
 | --- | ------ | -------------------------------------------------------------------------------------------------------------------- | ----- |
-| 7   | done   | `agents/webster-monitor.json` (Haiku 4.5) — analytics anomaly detection                                              | 1     |
+| 7   | done   | `agents/production/webster-monitor.json` (Haiku 4.5) — analytics anomaly detection                                   | 1     |
 | 8   | done   | 5 specialist critic specs: seo, brand-voice, fh-compliance, conversion, copy (all Sonnet 4.6) — all schema-valid     | 4     |
-| 9   | done   | `agents/webster-redesigner.json` (Opus 4.7) — synthesis + proposal generation                                        | 1     |
+| 9   | done   | `agents/production/webster-redesigner.json` (Opus 4.7) — synthesis + proposal generation                             | 1     |
 | 10  | done   | GitHub MCP integration — URL-based, vault-bound (`vault_ids`), no tokens in `user.message`                           | 3     |
 | 11  | done   | Environment config — `environments/webster-council-env.json` + `.id` registered                                      | 2     |
 | 12  | done   | Parallel fan-out via orchestrator → 6 parallel `/v1/sessions` calls (not `callable_agents`; that's research-preview) | 2     |
@@ -132,7 +132,7 @@ Key design calls (locked session 4 Phase 7, see ADR-0002):
 
 | #   | Status | Feature                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               | Hours |
 | --- | ------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----- |
-| 41a | done   | **Visual-reviewer agent spec** — `agents/webster-visual-reviewer.json` (Opus 4.7 tier). Inputs: preview URL, `history/<week>/proposal.md`, BEFORE URL. Outputs: `history/<week>/visual-review.md` with findings + embedded screenshot refs.                                                                                                                                                                                                                                                                                                           | 1     |
+| 41a | done   | **Visual-reviewer agent spec** — `agents/production/webster-visual-reviewer.json` (Opus 4.7 tier). Inputs: preview URL, `history/<week>/proposal.md`, BEFORE URL. Outputs: `history/<week>/visual-review.md` with findings + embedded screenshot refs.                                                                                                                                                                                                                                                                                                | 1     |
 | 41b | done   | **Browser-audit skill** — `skills/webster-browser-audit/SKILL.md` wraps `scripts/browser-audit.ts` for Playwright-headless when available, with fallback artifacts when unavailable. Capabilities: 3-breakpoint screenshot (375/768/1440), accessibility-tree text extraction, interaction recording, and console log capture.                                                                                                                                                                                                                        | 3     |
 | 41c | done   | **Proposal-intent verifier** — `scripts/proposal-intent-verifier.ts` reads each issue in `proposal.md` and verifies visible phrase presence in rendered accessibility text (not source grep). Catches content drops like session-4 "No more patient churn" regression; layout overflow is covered by browser-audit summaries.                                                                                                                                                                                                                         | 2     |
 | 41d | done   | **#39 integration pattern** — apply worker now runs the visual-reviewer gate after #39c and before #39d PR emission. It retries up to 3 iterations, records `visual_review` in `apply-log.json`, and forces draft/partial PR metadata on CRITICAL visual regressions.                                                                                                                                                                                                                                                                                 | 1     |
@@ -160,17 +160,17 @@ Added session 4 Phase 7. Autoresearch is **input to the next council run**, not
 
 Session 4 Phase 7 locked 9 architectural questions (Q1–Q9) — all resolved in `context/DOMAIN-MODEL.md`. Key locks: Q1 Managed Agent + orchestrator-owned memory (ADR-0001), Q2 explore-broadly cold-start + unified `history/memory.jsonl`, Q3 autonomous p<0.01 rollback, Q4 reward+gates 7-outcome matrix (ADR-0002), Q5 planner-requests-new-critic via L3 genealogy (additive-only), Q5.1 four-layer genealogy governance, Q6 skip-is-terminal + structured skip rows, Q7 Pair Alpha (SaaS + local service) substrate pair, Q8 per-experiment baselines + commit trailers (ADR-0002), Q9 4-week demo arc.
 
-| #   | Status | Feature                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  | Hours |
-| --- | ------ | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----- |
-| 50  | done   | **`agents/webster-planner.json`** (Opus 4.7 Managed Agent, Q1 ADR-0001) — registered via `POST /v1/agents`, invoked per-run via `/v1/sessions` + events + poll (pattern verified in `scripts/critic-genealogy.ts:440-556`). Reads marshaled memory context, outputs `plan.md` with `{classification, next_action, direction_hint, new_critic_request?, rationale}`. `next_action ∈ {promote_and_experiment, hold_baseline, revert_and_retry, explore_broadly}`. _Landed on `forge/task-feat-planner-agent-spec-v5` — PR #3 merged (2026-04-24)._         | 2     |
-| 51  | done   | **Memory substrate schema + append helper** (Q2) — `history/memory.jsonl` event log: `{ts, week, actor, event, refs{}, insight}` where `event ∈ {promote, rollback, skip, regression, gap-detected, verdict-ready}`. Append-only. Helper in orchestrator never touched by agents — orchestrator owns all I/O (ADR-0001). _Landed on `forge/task-feat-memory-substrate` — all 4 stories (MemoryEvent types, appendEvent, tailN+filter, unit tests)._                                                                                                      | 2     |
-| 52  | done   | **Orchestrator memory marshaling + planner invocation** (Q1, Q2) — new step in `prompts/second-wbs-session.md`: before critics, orchestrator reads `memory.jsonl` tail (last N events) + last 2 weeks' `verdict.json` + `monitor` anomaly report, concatenates to planner's user-message text (step 3 of the 5-step Managed Agent flow), polls until idle, extracts output, writes `history/<week>/plan.md`, appends one `verdict-ready` event row to `memory.jsonl`. _Landed on `forge/task-feat-orch-memory-planner-v2` — PR #6 merged (2026-04-24)._  | 3     |
-| 53  | done   | **Plan → council integration (additive-only)** (Q5) — critics + monitor + redesigner now receive `plan.md` body in initial `user.message` context with explicit additive-only/sovereignty language. Planner `new_critic_request` is extracted to `tmp/planner-new-critic-request-<week>.json` and passed into `scripts/critic-genealogy.ts --planner-request` as additive evidence, without bypassing dedup/cap/evidence gates.                                                                                                                          | 3     |
-| 54  | done   | **Cold-start explore-broadly mode** (Q2) — planner context now emits `direction_hint="broad exploration, baseline-only analytics"` when memory/verdict/monitor inputs are empty, and `appendColdStartOriginEvent()` writes the origin event row.                                                                                                                                                                                                                                                                                                         | 2     |
-| 55  | done   | **Genealogy governance layers 2–4** (Q5.1) — layer 1 is prompt-only (rubric in planner + redesigner instructions: "request only if existing critics cannot cover the concern"). Layer 2: orchestrator-side dedup — reject new-critic spec if ≥60% scope overlap with existing critic (embedding cosine). Layer 3: quarterly cap — max 3 new critics / 13 weeks, soft-override by operator. Layer 4: retire-on-idle — critic with 0 findings-promoted in 8 weeks is archived. _Landed on `forge/task-feat-genealogy-gov-v1` — PR #8 merged (2026-04-24)._ | 3     |
-| 56  | done   | **Skip-contract plumbing** (Q6) — apply-worker, critic-rerun gate, and visual-review gate now emit canonical structured skip rows to `history/<week>/skips.jsonl` and append skip events to `history/memory.jsonl` with reasons `{apply-fail, critic-veto, visual-veto}`. Skip is terminal and feeds next-week planner.                                                                                                                                                                                                                                  | 2     |
-| 57  | done   | **`scripts/seed-demo-arc.ts`** (Q9) — 4-week primary-substrate mock: 9 experiments + 1 genealogy spawn in W4. Hits 6/7 Q4 outcome lanes (fast-track, fallback, gate-win, archive-gate-fail, auto-rollback, hold). Idempotent; writes to `history/demo-arc/` without touching live history. _Landed on `forge/task-feat-seed-demo-arc-w3w4-v5` — all 4 stories (W1/W2/W3/W4 + genealogy) done, PR #5 merged (2026-04-24)._                                                                                                                                | 3     |
-| 58  | done   | **`scripts/seed-secondary-substrates.ts`** (Q7) — Pair Alpha mock: SaaS (B2B) + local service (B2C) synthetic single-file HTMLs + 2-cycle mock runs each (onboard + 2 weeks of experiments). Proves generalization beyond the primary substrate. Demo-safe (no e-commerce — private hold-out per operator decision). _Landed on `forge/task-feat-seed-pair-alpha-v1` — PR #7 merged (2026-04-24)._                                                                                                                                                       | 3     |
+| #   | Status | Feature                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     | Hours |
+| --- | ------ | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----- |
+| 50  | done   | **`agents/production/webster-planner.json`** (Opus 4.7 Managed Agent, Q1 ADR-0001) — registered via `POST /v1/agents`, invoked per-run via `/v1/sessions` + events + poll (pattern verified in `scripts/critic-genealogy.ts:440-556`). Reads marshaled memory context, outputs `plan.md` with `{classification, next_action, direction_hint, new_critic_request?, rationale}`. `next_action ∈ {promote_and_experiment, hold_baseline, revert_and_retry, explore_broadly}`. _Landed on `forge/task-feat-planner-agent-spec-v5` — PR #3 merged (2026-04-24)._ | 2     |
+| 51  | done   | **Memory substrate schema + append helper** (Q2) — `history/memory.jsonl` event log: `{ts, week, actor, event, refs{}, insight}` where `event ∈ {promote, rollback, skip, regression, gap-detected, verdict-ready}`. Append-only. Helper in orchestrator never touched by agents — orchestrator owns all I/O (ADR-0001). _Landed on `forge/task-feat-memory-substrate` — all 4 stories (MemoryEvent types, appendEvent, tailN+filter, unit tests)._                                                                                                         | 2     |
+| 52  | done   | **Orchestrator memory marshaling + planner invocation** (Q1, Q2) — new step in `prompts/second-wbs-session.md`: before critics, orchestrator reads `memory.jsonl` tail (last N events) + last 2 weeks' `verdict.json` + `monitor` anomaly report, concatenates to planner's user-message text (step 3 of the 5-step Managed Agent flow), polls until idle, extracts output, writes `history/<week>/plan.md`, appends one `verdict-ready` event row to `memory.jsonl`. _Landed on `forge/task-feat-orch-memory-planner-v2` — PR #6 merged (2026-04-24)._     | 3     |
+| 53  | done   | **Plan → council integration (additive-only)** (Q5) — critics + monitor + redesigner now receive `plan.md` body in initial `user.message` context with explicit additive-only/sovereignty language. Planner `new_critic_request` is extracted to `tmp/planner-new-critic-request-<week>.json` and passed into `scripts/critic-genealogy.ts --planner-request` as additive evidence, without bypassing dedup/cap/evidence gates.                                                                                                                             | 3     |
+| 54  | done   | **Cold-start explore-broadly mode** (Q2) — planner context now emits `direction_hint="broad exploration, baseline-only analytics"` when memory/verdict/monitor inputs are empty, and `appendColdStartOriginEvent()` writes the origin event row.                                                                                                                                                                                                                                                                                                            | 2     |
+| 55  | done   | **Genealogy governance layers 2–4** (Q5.1) — layer 1 is prompt-only (rubric in planner + redesigner instructions: "request only if existing critics cannot cover the concern"). Layer 2: orchestrator-side dedup — reject new-critic spec if ≥60% scope overlap with existing critic (embedding cosine). Layer 3: quarterly cap — max 3 new critics / 13 weeks, soft-override by operator. Layer 4: retire-on-idle — critic with 0 findings-promoted in 8 weeks is archived. _Landed on `forge/task-feat-genealogy-gov-v1` — PR #8 merged (2026-04-24)._    | 3     |
+| 56  | done   | **Skip-contract plumbing** (Q6) — apply-worker, critic-rerun gate, and visual-review gate now emit canonical structured skip rows to `history/<week>/skips.jsonl` and append skip events to `history/memory.jsonl` with reasons `{apply-fail, critic-veto, visual-veto}`. Skip is terminal and feeds next-week planner.                                                                                                                                                                                                                                     | 2     |
+| 57  | done   | **`scripts/seed-demo-arc.ts`** (Q9) — 4-week primary-substrate mock: 9 experiments + 1 genealogy spawn in W4. Hits 6/7 Q4 outcome lanes (fast-track, fallback, gate-win, archive-gate-fail, auto-rollback, hold). Idempotent; writes to `history/demo-arc/` without touching live history. _Landed on `forge/task-feat-seed-demo-arc-w3w4-v5` — all 4 stories (W1/W2/W3/W4 + genealogy) done, PR #5 merged (2026-04-24)._                                                                                                                                   | 3     |
+| 58  | done   | **`scripts/seed-secondary-substrates.ts`** (Q7) — Pair Alpha mock: SaaS (B2B) + local service (B2C) synthetic single-file HTMLs + 2-cycle mock runs each (onboard + 2 weeks of experiments). Proves generalization beyond the primary substrate. Demo-safe (no e-commerce — private hold-out per operator decision). _Landed on `forge/task-feat-seed-pair-alpha-v1` — PR #7 merged (2026-04-24)._                                                                                                                                                          | 3     |
 
 ## Totals (historical — initial plan)