@event4u/agent-config 2.12.0 → 2.14.0

package/.agent-src/rules/copilot-routing.md

@@ -0,0 +1,19 @@
+---
+type: "auto"
+tier: "3"
+description: "Configuring GitHub Copilot — copilot-instructions.md, PR-review comment patterns, suggestion behavior — route to the copilot-config skill"
+source: package
+triggers:
+  - keyword: "copilot"
+  - phrase: "copilot-instructions"
+  - phrase: "copilot pr review"
+routes_to:
+  - "skill:copilot-config"
+---
+
+# Copilot Routing
+
+**Iron Law.** Tuning the GitHub Copilot AI assistant itself (instructions, PR-review patterns, suggestion behavior) → load the `copilot-config` skill, not `devcontainer` (which covers the dev environment Copilot runs inside).
+
+Body migrated to `skill:copilot-config`. Disambiguates the copilot-config ↔ devcontainer cluster head per [`adr-architectural-consensus-mechanism`](../docs/contracts/adr-architectural-consensus-mechanism.md).
+Trigger-set above activates this routing under the `balanced` and `full` profiles.

package/.agent-src/rules/devcontainer-routing.md

@@ -0,0 +1,20 @@
+---
+type: "auto"
+tier: "3"
+description: "Wiring DevContainers or GitHub Codespaces — devcontainer.json, images, VS Code features, port forwarding — route to the devcontainer skill"
+source: package
+triggers:
+  - keyword: "devcontainer"
+  - keyword: "codespaces"
+  - keyword: "codespace"
+  - phrase: "devcontainer.json"
+routes_to:
+  - "skill:devcontainer"
+---
+
+# Devcontainer Routing
+
+**Iron Law.** Wiring the dev environment itself (DevContainers, Codespaces, `devcontainer.json`, VS Code features) → load the `devcontainer` skill, not `copilot-config` (which tunes the Copilot AI on top).
+
+Body migrated to `skill:devcontainer`. Disambiguates the devcontainer ↔ copilot-config cluster head per [`adr-architectural-consensus-mechanism`](../docs/contracts/adr-architectural-consensus-mechanism.md).
+Trigger-set above activates this routing under the `balanced` and `full` profiles.

package/.agent-src/rules/external-reference-deep-dive.md

@@ -1,7 +1,7 @@
 ---
 type: "auto"
 tier: "2b"
-description: "When the user names an external repo, file, URL, or artifact as a reference — fetch the actual tree and inspect, never summarize from README or metadata"
+description: "User names an external repo, file, URL, or artifact as reference — fetch the actual tree and inspect, never summarize from README or metadata"
 alwaysApply: false
 source: package
 triggers:

package/.agent-src/rules/fast-path-marker-visibility.md

@@ -0,0 +1,38 @@
+---
+type: "auto"
+tier: "1"
+description: "Low-impact council fast-path dispatch — host agent MUST surface the transparency marker verbatim in the reply opening; never swallow or paraphrase it."
+source: package
+triggers:
+  - keyword: "low-impact council"
+  - keyword: "fast-path"
+  - keyword: "Resolved via low-impact council"
+  - keyword: "low_impact"
+  - intent: "low-impact council dispatch"
+validator_ignore:
+  - type: "substring"
+    pattern: ".agent-src.uncompressed/"
+    reason: "Compressor injects a back-pointer to the uncompressed source for full failure-modes detail."
+---
+
+# Fast-Path Marker Visibility
+
+## The Iron Law
+
+```
+EVERY LOW-IMPACT COUNCIL FAST-PATH REPLY OPENS WITH THE EXACT MARKER.
+NEVER PARAPHRASE. NEVER SWALLOW. NEVER SUBSTITUTE THE AGENT'S OWN VERDICT.
+```
+
+Markers (from `scripts/ai_council/low_impact.py`):
+
+- **Resolved** — `> Resolved via low-impact council fast-path: <verdict>.`
+- **Unavailable** — `> Low-impact council unavailable (no opted-in members) — escalating to user.`
+- **Split** — `> Low-impact council split — escalating to user (<m1>: X / <m2>: Y):`
+- **Aborted** — `> Low-impact council aborted (token cap) — escalating to user:`
+
+Verbatim = first non-whitespace line, English (no translation), no emoji prefix, no merged numbered-options. Marker is the only audit signal that distinguishes fast-path from local deliberation. See `.agent-src.uncompressed/rules/fast-path-marker-visibility.md` for full failure modes.
+
+Scope: `low_impact` class only. `high_impact` and `user_required` never reach fast-path.
+
+See: [`ai-council-config § Low-impact council opt-in`](../docs/contracts/ai-council-config.md#low-impact-council-opt-in), [`direct-answers`](direct-answers.md) (invented-facts Iron Law kin).

package/.agent-src/rules/laravel-routing.md

@@ -0,0 +1,20 @@
+---
+type: "auto"
+tier: "3"
+description: "When writing or reviewing Laravel code — controllers, Eloquent, Artisan, jobs, events, policies — route to the laravel skill"
+source: package
+triggers:
+  - keyword: "laravel"
+  - keyword: "artisan"
+  - keyword: "eloquent"
+  - keyword: "FormRequest"
+routes_to:
+  - "skill:laravel"
+---
+
+# Laravel Routing
+
+**Iron Law.** Laravel-flavoured PHP (Eloquent, Artisan, FormRequest, jobs, events, policies) → load the `laravel` skill, not `symfony-workflow` and not `php-coder`.
+
+Body migrated to `skill:laravel`. Disambiguates the laravel ↔ symfony-workflow cluster head per [`adr-architectural-consensus-mechanism`](../docs/contracts/adr-architectural-consensus-mechanism.md).
+Trigger-set above activates this routing under the `balanced` and `full` profiles.

package/.agent-src/rules/low-impact-corpus-privacy-floor.md

@@ -0,0 +1,74 @@
+---
+type: "auto"
+tier: "1"
+description: "Writing, editing, or upstreaming entries in `agents/low-impact-decisions.md` — non-bypassable privacy floor for the learning corpus."
+source: package
+triggers:
+  - path_prefix: "agents/low-impact-decisions"
+  - keyword: "low-impact-decisions"
+  - keyword: "low-impact corpus"
+  - keyword: "learn-low-impact"
+---
+
+# Low-Impact Corpus — Privacy Floor
+
+## Iron Law
+
+```
+NO ENTRY LEAVES THE PROJECT REPO UNTIL THE REDACTOR CLEARS IT.
+NO SECRETS. NO EMAILS. NO PROJECT PATHS. NO CUSTOMER NAMES.
+NO INTERNAL HOSTNAMES. NO MONEY. NO BUSINESS SQL. NO LONG CODE.
+```
+
+Redactor lives in `scripts/ai_council/redact_low_impact_entry.py`
+and runs at **both** gates:
+
+1. **Write gate** — every intake append to
+   `agents/low-impact-decisions.md` (Phase 12 § Step 2).
+2. **Upstream gate** — every `/memory learn-low-impact` PR draft,
+   before diff leaves repo (Phase 12 § Step 5).
+
+Failure at either gate refuses the operation, surfaces offending
+pattern, asks user to rephrase. Redactor never auto-rewrites — silent
+rewriting is soft gate, this is hard.
+
+## Forbidden-content classes (8)
+
+| # | Class | Pattern source |
+|---|---|---|
+| 1 | Secrets | raw-key prefixes from `scripts/ai_council/config._RAW_KEY_PREFIXES` + inline `api_key:` shape |
+| 2 | Emails | RFC-5322-ish |
+| 3 | Project-rooted paths | `/Users/`, `/home/`, `/opt/`, `/private/`, drive letters, configured `repo_root` |
+| 4 | Customer / tenant names | caller-supplied list; generic placeholders (`<customer>`, `<tenant>`, `<account>`, `<user>`) survive |
+| 5 | Internal hostnames | `*.internal`, `*.local`, caller-supplied private domains |
+| 6 | Monetary amounts | `$1,234`, `€500`, `USD 1000` shapes |
+| 7 | Business-context SQL identifiers | caller-supplied table / column list |
+| 8 | Inline code excerpts > 40 chars | backtick-fenced runs |
+
+## Locked target types
+
+- `agents/low-impact-decisions.md` — project-local corpus.
+- `data/low-impact-decisions-seed.md` (agent-config package) —
+  upstream seed shipped with the package.
+
+## When to invoke
+
+- Host agent just received user intake trigger (see
+  `scripts/ai_council/low_impact_intake.TRIGGER_PHRASES`).
+- Host agent about to write entry to corpus or open `/memory learn-low-impact` PR.
+
+## What to surface on refusal
+
+One-line marker:
+
+```
+> Low-impact corpus refused — <category>: <snippet…>. Rephrase or skip.
+```
+
+Then agent stops intake/upstream flow. No silent retry, no auto-rewrite.
+
+## See also
+
+- `scripts/ai_council/redact_low_impact_entry.py` — the redactor.
+- `scripts/ai_council/low_impact_intake.py` — write-gate caller.
+- `agents/low-impact-decisions.md` — corpus + Anti-Examples list.

package/.agent-src/rules/symfony-routing.md

@@ -0,0 +1,20 @@
+---
+type: "auto"
+tier: "3"
+description: "Writing or reviewing Symfony code — DI, bundles, Doctrine, Messenger, Security voters, console commands — route to the symfony-workflow skill"
+source: package
+triggers:
+  - keyword: "symfony"
+  - keyword: "doctrine"
+  - keyword: "twig"
+  - keyword: "messenger"
+routes_to:
+  - "skill:symfony-workflow"
+---
+
+# Symfony Routing
+
+**Iron Law.** Symfony-flavoured PHP (DI container, bundles, Doctrine entities, Messenger, Security voters, console commands) → load the `symfony-workflow` skill, not `laravel` and not `php-coder`.
+
+Body migrated to `skill:symfony-workflow`. Disambiguates the laravel ↔ symfony-workflow cluster head per [`adr-architectural-consensus-mechanism`](../docs/contracts/adr-architectural-consensus-mechanism.md).
+Trigger-set above activates this routing under the `balanced` and `full` profiles.

package/.agent-src/skills/ai-council/SKILL.md

@@ -5,6 +5,8 @@ source: package
 domain: process
 ---
 
+<!-- cloud_safe: degrade -->
+
 > **Experimental.** AI Council is not yet validated by external users. API costs apply per consultation.
 
 # ai-council
@@ -30,6 +32,45 @@ Do NOT use when:
 * The user has not configured any council member → state that and stop;
   do not silently fall back to anything.
 
+## When NOT to invoke — necessity self-check
+
+Phase 6 necessity classifier (see
+[`ai-council-config § Necessity classifier`](../../../docs/contracts/ai-council-config.md))
+runs as pre-flight gate inside CLI, skips council when prompt looks
+like routine work. Route around it BEFORE gate fires so user never pays
+classifier-pause cost on request that obviously did not need council.
+
+Skip council, stay in-session for:
+
+* **Bugfix shape** — stack trace, error, crash, failing test, "broken",
+  regression. Use `systematic-debugging` or `bug-investigate`.
+* **Syntax / format / lint** — `typo`, `formatting`, `lint`, `indent`,
+  `import order`, simple rename. Use language skill directly
+  (`php-coder`, `eloquent`, `nextjs-patterns`).
+* **Single-file implementation** — "this function", "this method",
+  "this file", "one-liner", "small change", "add a getter". Use
+  language skill directly.
+* **Documentation lookup** — "what is X", "how does Y work", "example
+  of Z", "syntax of W". Use `codebase-retrieval` or docs skill, never
+  council.
+
+Invoke council when:
+
+* **Architectural / structural** — system boundaries, coupling,
+  refactor strategy, migration plan, rewrite vs redesign.
+* **Multi-axis trade-off** — stakeholders disagree; competing
+  alternatives need weighing; "pros and cons" is the actual ask.
+* **Strategic / direction** — "should we …", "shall we …", roadmap
+  shape, long-term technical direction.
+* **Explicit ambiguity** — user wrote "unsure / uncertain / ambiguous
+  / second opinion / sanity check".
+
+Agent orchestration MUST call `council_cli` with `--invocation agent`
+so gate can skip silently on routine requests. User-typed `/council`
+keeps default (`--invocation user_explicit`); user gets educational
+message + `--proceed-anyway` override path. Mode `block` ignores
+`--proceed-anyway` by design — cost-strict opt-in.
+
 ## Goal
 
 Bring in **independent** external models to critique a project
@@ -89,8 +130,9 @@ travel changes.
 
 | Mode | Client | Billable | Transport | Status |
 |---|---|---|---|---|
-| `api` | `AnthropicClient` / `OpenAIClient` | yes | provider SDK + key from `~/.event4u/agent-config/<provider>.key` (legacy `~/.config/agent-config/<provider>.key` read as fallback) | shipped |
+| `api` | `AnthropicClient` / `OpenAIClient` / `GeminiClient` / `XAIClient` / `PerplexityClient` | yes | provider SDK + key from `~/.event4u/agent-config/<provider>.key` (legacy `~/.config/agent-config/<provider>.key` read as fallback) | shipped |
 | `manual` | `ManualClient` | no | `stdout` (prompt block) + `stdin` (user pastes the web-UI reply, terminated by a line containing only `END`) | shipped (Phase 2b) |
+| `cli` | `AnthropicCliClient` / `OpenAICliClient` / `GeminiCliClient` | no (subscription-authed) | local subprocess against the vendor CLI (`claude`, `codex`, `gemini`); auth delegated to the CLI's own session, no API key in this process | shipped (anthropic/openai/gemini · Phase 3) |
 
 Resolution lives in `scripts/ai_council/modes.py`:
 `resolve_mode(name, invocation_mode, member_settings, global_mode)`
@@ -118,16 +160,79 @@ thread** (no system prompt repetition). `2` records the round and
 moves to the next member. `3` returns `error="manual_aborted"` for
 that member and the orchestrator stops the fan-out.
 
+### CLI-mode UX
+
+`mode: cli` runs the council through the vendor's local CLI
+instead of the API. Auth is delegated — user logs into each CLI
+once (`claude login`, `codex login`, `gemini`), orchestrator
+inherits the subscription. No API key in this process.
+`billable=False` → cost gate bypassed; the local
+`cli_call_budget.max_calls_per_day.<provider>` quota (state at
+`~/.event4u/agent-config/cli-calls.json`, daily UTC reset) is the
+only per-day brake.
+
+Three vendor CLIs wired:
+
+- **Anthropic / Claude** — invokes `claude --print --output-format json`,
+  parses standard envelope (`result` + `usage` + `session_id` +
+  `total_cost_usd`). Token counts and reported cost survive to
+  `metadata` for audit.
+
+  ```yaml
+  members:
+    anthropic:
+      enabled: true
+      mode: cli
+      model: claude-sonnet-4-5
+  ```
+
+- **OpenAI / Codex** — invokes `codex exec --json`, walks the
+  newline-delimited JSON event stream, pulls text from
+  `item.completed` and tokens from `turn.completed`. Session id
+  preserved.
+
+  ```yaml
+  members:
+    openai:
+      enabled: true
+      mode: cli
+      model: gpt-5
+  ```
+
+- **Google / Gemini** — invokes `gemini --output-format json` with
+  prompt piped on stdin, parses `response` + `stats.models.<m>.tokens`
+  envelope. OAuth consent must be granted once interactively before
+  the CLI is usable from a non-interactive shell.
+
+  ```yaml
+  members:
+    gemini:
+      enabled: true
+      mode: cli
+      model: gemini-2.5-pro
+  ```
+
+Auth-failure stderr from any vendor CLI surfaces as
+`error="auth_expired"` with the original stderr tail in
+`metadata.stderr_tail` so the user knows to re-login. Missing
+binary at construction time fails fast with `CouncilDisabledError`
+naming the binary and the YAML override path — never silently
+substitutes.
+
+`xai` + `perplexity` accept `mode: cli` from Phase 4 onward, but
+their community CLIs DO consume the API key and DO NOT bypass
+per-token billing — contract doc warns explicitly.
+
 ### Cost-gate bypass for non-billable members
 
 `ExternalAIClient.billable` is the contract. Clients with
-`billable=False` (`ManualClient`) bypass the cost gate entirely —
-the orchestrator skips the
-projection check, the `on_overrun` callback, and the USD-budget
-short-circuit for that member, but still records the response's
-token counts (from the manual-paste length heuristic or the
-provider's reply, when available) for observability. Mixed runs
-(one manual + one api) gate only the api members.
+`billable=False` (`ManualClient`, `AnthropicCliClient`,
+`OpenAICliClient`, `GeminiCliClient`) bypass the cost gate entirely —
+orchestrator skips the projection check, the `on_overrun` callback,
+and the USD-budget short-circuit for that member, but still records
+the response's token counts (from the manual-paste length heuristic
+or the provider's reply, when available) for observability. Mixed
+runs (one cli + one api) gate only the api members.
 
 ## Degradation modes
 
@@ -266,6 +371,16 @@ matching `road-to-<topic-slug>` roadmap under `agents/roadmaps/`).
   + gpt-4o, YYYY-MM-DD) reviewed N candidate strategies; converged
   on …`).
 
+### Exempt
+
+- `agents/audit-*/` — historical audit bundles. Canonical council
+  dirs are gitignored; audit bundles are tracked, cohesive narratives
+  that may include council artefacts as part of their evidence trail
+  (e.g. `audit-2026-05-14-north-star/` bundles its triggering
+  question, raw responses, and synthesis alongside the audit's
+  findings). The layout linter (`scripts/check_council_layout.py`)
+  skips these directories.
+
 `scripts/check_council_layout.py` is the mechanical check for the
 output path convention — wire it into the package's CI pipeline so
 violations break the build.
@@ -292,6 +407,65 @@ NEVER ships the host verdict as council output. Provider attribution
 stays visible in the per-member sections; host verdicts stay
 attributed to the host.
 
+## Synthesis templates (lens-aware)
+
+The **Convergence / Divergence** slot in `council:render` output is
+populated with a lens-specific synthesis prompt. The host agent reads
+this prompt and writes the summary in the shape it dictates. The five
+templates live in `scripts/ai_council/prompts.py::_SYNTHESIS_TABLE`
+and are exposed via `synthesis_template(mode)`.
+
+**R4 Q4 split** — decision lenses get a structured Karpathy-style
+template; creative lenses stay open-ended prose (bare slot):
+
+| Lens | Class | Synthesis sections |
+|---|---|---|
+| `default` | decision | Agreement · Clashes · Blind spots · Recommendation · Next step |
+| `pr` | decision | Consensus · Conflicts · Must-fix before merge · Recommendation |
+| `analysis` | decision | Top-10 by consensus · Supporting · Outliers |
+| `design` | creative | *(no template — open prose passthrough)* |
+| `optimize` | creative | *(no template — open prose passthrough)* |
+
+Input modes (`prompt` / `roadmap` / `diff` / `files`) inherit the
+`default` decision template — they are bundling shapes, not lenses.
+
+**Source citations:**
+* Template shape — Round 2 council verdict
+  (`agents/council-sessions/2026-05-14-ai-council-redesign/round-1.md`,
+  Opus's lens-adaptive synthesis proposal).
+* Decision/creative split — Round 4 Q4 verdict
+  (`agents/council-sessions/2026-05-14-ai-council-redesign/round-3-responses.json`).
+  R4 reframed `optimize` as creative because perf trade-offs resist
+  pre-templated shape — Performance-wins / Trade-offs /
+  Implementation-order is too rigid for the variety of optimize-lens
+  artefacts. R4 reframed `design` for the same reason — design
+  critiques are inherently narrative.
+
+### `--prose-synthesis` escape hatch (R4 Q4)
+
+Both `council:run` and `council:render` accept symmetric escape-hatch
+flags on top of the lens table:
+
+* `--prose-synthesis` — force creative-lens passthrough (bare slot)
+  regardless of lens. Use when the user on `default`/`pr`/`analysis`
+  prefers a narrative recommendation over the structured template.
+* `--no-prose-synthesis` — force the `default` structured template
+  even on a creative lens. Use when a `design` or `optimize` artefact
+  has a one-shot need for Karpathy-style structure.
+
+The flag is mutually exclusive — picking one disables the other on
+the same invocation. When `council:run` records either flag in the
+output JSON, `council:render` honours it unless the renderer is
+called with an explicit flag of its own.
+
+### Renderer lens resolution
+
+`council:render <responses.json>` resolves the active lens in this
+order: explicit `--prompt-mode` flag > `prompt_mode` field in the
+payload > `mode` field in the payload > `None` (default decision
+template). The `--prose-synthesis` / `--no-prose-synthesis` flag
+overrides the table regardless of how the lens resolved.
+
 ## Do NOT
 
 - Do NOT paraphrase council output into the host agent's voice — strip
@@ -435,8 +609,8 @@ prompt as `<original artefact> + <prior round, anonymised>` so each
 member can refine, agree, or push back on the previous critique
 without seeing which provider produced which point.
 
-The default round count comes from `ai_council.min_rounds` in
-`.agent-settings.yml` (default `2` so members critique each other
+The default round count comes from `defaults.min_rounds` in
+`agents/.ai-council.yml` (default `2` so members critique each other
 at least once before convergence). The host agent does **not** ask
 "how many rounds?" when the requested count is `<= min_rounds` —
 the settings owner already made that decision. Ask only when a
@@ -516,6 +690,202 @@ internal "more feedback" follow-up loop (1 / 2 / 3 menu) is
 **inside** a single member's chat thread and is orthogonal to the
 orchestrator-level rounds.
 
+### `/council debate` sub-command (progressive disclosure)
+
+`/council debate <artefact> [--rounds N] [--continue-as-debate <session>]`
+runs an **interactive** multi-round critique with a confirmation gate
+between every round so the user can stop the spend at any point.
+
+| Property | Behaviour |
+|---|---|
+| Round flow | Same orchestrator as `rounds:N` (`run_debate()`), but each round prints its responses then pauses on a y/n prompt before launching the next round. |
+| Cost gate | After every round the CLI prints `Spent so far: $X · Next round: ~$Y · Cap: $Z`. `n` exits cleanly with partial results; `y` continues. |
+| Hard cap | If the projected next-round cost would breach `max_total_usd`, `run_debate()` raises `DebateCapExceeded` and the CLI exits with the partial transcript. No silent overrun. |
+| `--continue-as-debate` | Seeds round 1 from an existing `/council default` (or analysis lens) session. No round-1 API calls are billed; round 2+ run normally. Member list must match. |
+| Session files | One file per round under `agents/council-sessions/<slug>/debate-round-NN.md`. |
+| Anonymisation | Identical to `rounds:N`. The continue-as-debate path also anonymises the seeded round-1 responses when building the round-2 prompt. |
+
+Use this when the artefact is genuinely contentious and the user
+wants to control depth interactively. For a fire-and-forget
+multi-round run, prefer `consult(..., rounds=N)` or `--rounds N` on
+`/council default`.
+
+
+## Karpathy peer-review (opt-in)
+
+After the final deliberation round, an optional **anonymous peer-review
+pass** lets each member critique the *other* members' responses for
+blind spots before synthesis. Inspired by Andrej Karpathy's "ask the
+strongest models to review each other anonymously" pattern; see his
+[talks / threads on inter-model critique](https://karpathy.ai/) and the
+internal verdict in
+`agents/council-sessions/2026-05-14-ai-council-redesign/round-2.md`
+(R2 split: one approve-as-flag, one reject-as-default → opt-in only).
+
+Pipeline order when every feature is active:
+
+```
+deliberation rounds → peer-review → consensus-scoring → synthesis
+```
+
+Activation — two equivalent paths:
+
+* CLI: `--peer-review` on `council:estimate` or `council:run`.
+* Config: `ai_council.peer_review.enabled: true` in
+  `agents/.ai-council.yml`. Default is `false`.
+
+Mechanics:
+
+1. The final deliberation round's outputs are anonymised into
+   `Response-A`, `Response-B`, … in stable input order. Provider /
+   model identity is stripped (Iron-Law neutrality holds); empty or
+   errored deliberation responses are skipped.
+2. Each member receives an N−1 view (its own response filtered out)
+   plus the Karpathy prompt: *strongest response*, *weakest blind
+   spot*, *what did everyone miss*, *refinement*.
+3. The N critiques flow back into synthesis through a
+   "Peer-Review-Surfaced Blind Spots" addendum on the lens template.
+4. **Advisor preserve-persona (R4 Q3, hard-coded):** when the
+   deliberation was an advisor-mode run (Phase 6), anonymisation
+   strips provider identity but **preserves the advisor persona
+   label**. Peer-review renders as `Response A (Contrarian)`, never
+   `Response A (Anthropic Opus)`. Plain-member runs strip identity
+   entirely.
+
+Cost — adds exactly N billable calls (one per member) at the same
+per-call cost as a deliberation call. The `council:estimate` table
+surfaces the delta as a `+peer-review: +N calls (~+$X)` row.
+
+Needs ≥ 2 distinct deliberation outputs; below that the round is a
+no-op and nothing extra is billed. Self-review is structurally
+impossible — a member never sees its own response.
+
+## Thinking-style advisors (replace-mode)
+
+Phase 6 introduces five **advisor personas** the council adopts in
+*replace-mode*: an enabled advisor substitutes its bound member's
+plain call with the same provider running the advisor's persona
+prompt. Total call count stays the same — only the system prompt
+swaps.
+
+| Advisor | Default bound member | Focus |
+|---|---|---|
+| **Contrarian** | `anthropic` | strongest counterargument, hidden assumptions |
+| **First-Principles** | `anthropic` | strip metaphor, derive from physics / math / cost |
+| **Expansionist** | `openai` | adjacent opportunities, second-order effects |
+| **Outsider** | `openai` | naive-but-sharp questions, beginner's-mind probes |
+| **Executor** | `anthropic` | what ships this quarter, what blocks delivery |
+
+Activation — edit `agents/.ai-council.yml` and flip the advisor's
+`enabled: true`. Optional `model: <name>` overrides the bound
+member's default model. An advisor referencing a disabled member
+fails closed at config load — never silently skipped.
+
+```yaml
+advisors:
+  contrarian:
+    enabled: true        # ← swap anthropic's plain call for contrarian
+    member: anthropic
+    # model: claude-opus-4   # optional pin
+```
+
+`council:estimate` surfaces every active swap on a dedicated line
+above the cost table:
+
+```
+council:estimate · mode=prompt · members=2 (billable=2)
+  advisor: Contrarian on anthropic via claude-sonnet-4-5
+anthropic/claude-sonnet-4-5: ~991 in + 256 out  =  $0.0068
+openai/gpt-4o: ~208 in + 256 out  =  $0.0031
+```
+
+Cost-bounded — replace-mode never adds calls. The persona prompt is
+larger than plain (~1k extra input tokens per swap); output tokens
+and call count are unaffected. Peer-review preserves the advisor
+label while stripping provider identity (`Response A (Contrarian)`).
+Two enabled advisors on the same member is a config error.
+
+## Decision-replay artefact (Phase 9, audit trail)
+
+Every session that runs consensus scoring drops a `decision-replay.md`
+next to `responses.json`. Pure projection of consensus + final-round
+member texts — **no extra model calls, no extra spend**. Surfaces per
+top finding: verdict band (Strong/Moderate/Weak), evidence-quality
+(H/M/L), agree/dissent split, one key argument per member.
+
+Two render modes — **Full** (per-member arguments attributed to
+`provider:model`) and **Redacted** (verdict + evidence-quality + counts
+only). Toggles: `ai_council.decision_replay.{enabled,
+include_member_arguments}` global; `ai_council.lenses.<lens>.decision_replay.*`
+per-lens override.
+
+CLI — written automatically by `council run` on lenses that score
+consensus. `council replay <responses.json>` re-renders from a saved
+session; `--redact-member-arguments` / `--include-member-arguments`
+flip the view independent of config (share redacted variant of an
+already-paid run).
+
+## Lightweight-QA fast-path (Phase 11)
+
+Low-impact questions from Phase 10's impact router → restricted fast-path
+in place of full debate. Trade-off explicit: **1 round · ≤2 members ·
+$0.05/answer · 2500 tokens**. No advisors, no peer-review, no consensus
+scoring — quick answer + transparency marker, not deliberation.
+
+### Iron Law
+
+`high_impact` and `user_required` **never** route to fast-path,
+regardless of config. Schema validation rejects override. Fast-path
+activates only when:
+
+1. `ai_council.enabled: true` AND
+2. `decision_resolution.low_impact.mode: council` AND
+3. ≥1 member has `participate_low_impact: true` (default `false` —
+   explicit opt-in per member).
+
+Default `low_impact` route = **`agent`** — nothing reaches council
+without explicit two-knob opt-in (flip class → `council` *and* mark
+≥1 member `participate_low_impact: true`). Worked YAML, validation,
+unavailable-marker contract → [`ai-council-config § Low-impact council opt-in`](../../../docs/contracts/ai-council-config.md#low-impact-council-opt-in).
+
+### Output marker (always surfaced)
+
+* **Resolved** — `> Resolved via low-impact council (anthropic): <answer>`
+* **Split** — `> Low-impact council split — escalating to user (anthropic: X / openai: Y):`
+* **Aborted** — `> Low-impact council aborted (token cap) — escalating to user:`
+
+Marker mandatory — agent never silently substitutes fast-path verdict
+for its own answer.
+
+### Session artefact
+
+Every fast-path attempt appends one line to
+`agents/council-sessions/<date>-<slug>/low-impact-resolutions.md`:
+
+```
+2025-05-14T10:00:00Z | resolved | members=2/2 | members(anthropic, openai) cost=$0.0034 | Q=Service vs Repository for this read path?
+```
+
+Append-only, one line per resolution. Parser tolerates free-form
+headers around canonical lines.
+
+### `council replay --low-impact-stats`
+
+Re-projection of session log → summary block:
+
+```
+$ council replay agents/council-sessions/2025-05-14-foo/responses.json --low-impact-stats
+# Low-impact fast-path · session summary
+
+- attempts: 4
+- status: aborted=1 · resolved=2 · split=1
+- members: anthropic=4 · openai=3
+- total cost: $0.0096
+```
+
+No model calls — pure markdown parse. Returns 0 when session has no
+fast-path entries (clean session is not an error).
+
 ## See also
 
 - `/council` command — the user-facing entry point.
@@ -523,6 +893,14 @@ orchestrator-level rounds.
   network, no spend, but no diversity of weights either).
 - `scripts/ai_council/prompts.py` — neutrality preamble + per-mode
   system prompts.
+- `scripts/ai_council/advisors.py` — replace-mode planning + persona
+  resolution.
 - `scripts/ai_council/bundler.py` — redaction pattern set + size
   guard.
 - `docs/customization.md` § `ai_council.*` — settings reference.
+- `docs/contracts/ai-council-config.md` § advisors — schema + precedence
+  contract.
+- `docs/contracts/ai-council-config.md` § Decision-replay artefact —
+  Phase 9 audit trail contract + redaction modes.
+- `scripts/ai_council/replay.py` — pure projection renderer (no model
+  calls).

package/.agent-src/skills/copilot-config/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: copilot-config
-description: "Use when configuring GitHub Copilot — copilot-instructions.md, PR review patterns, output optimization — even when the user just says 'tune Copilot' or 'why is Copilot commenting on X'."
+description: "Tune the GitHub Copilot AI — `copilot-instructions.md`, PR-review patterns, suggestion behavior, output verbosity. NOT for dev-environment setup (use `devcontainer`)."
 source: package
 domain: process
 ---