28 changed files with 332 additions and 1596 deletions
--- a/.claude-plugin/plugin.json
+++ b/.claude-plugin/plugin.json
@ -1,11 +1,11 @@
 {
  "name": "voyage",
  "description": "Voyage — brief, research, plan, execute, review, continue. Contract-driven Claude Code pipeline. /trekbrief, /trekplan, and /trekreview each end by building a self-contained operator-annotation HTML (scripts/annotate.mjs, modelled on claude-code-100x): select text or click any element, pick intent (Fiks/Endre/Spørsmål), write comment, copy structured prompt, paste back, Claude revises the .md.",
-  "version": "5.10.1",
+  "version": "5.10.0",
  "author": {
    "name": "Kjell Tore Guttormsen"
  },
-  "homepage": "https://git.fromaitochitta.com/open/voyage",
+  "homepage": "https://git.fromaitochitta.com/open/ktg-plugin-marketplace/src/branch/main/plugins/voyage",
  "repository": "https://git.fromaitochitta.com/open/ktg-plugin-marketplace.git",
  "license": "MIT",
  "keywords": [
--- a/.gitignore
+++ b/.gitignore
@ -48,6 +48,3 @@ NEXT-SESSION-PROMPT*.local.md
 # broad `git add -A docs`, which needed a history rewrite to undo. Ignoring the
 # type is cheaper than remembering not to stage it.
 docs/*.pdf
 # Page renders of those same documents (SLDC-AI-1_1.png etc). Narrow on purpose:
 # legitimate plugin assets may still be .png under docs/.
 docs/SLDC-AI*.png
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@ -4,111 +4,6 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 ## v5.10.1 — 2026-09-03 — gemini-bridge dropped; fail-closed review verdicts
 ### Removed
 - **`gemini-bridge` is dropped from the pipeline (operator directive).** The
  agent file is deleted and every live surface that could still steer a run
  toward it is cleared: `commands/trekresearch.md` (the Bridge-agent launch
  block, the `--local` help text, the high-effort always-on pair, the stats
  record, the graceful-degradation list), `commands/trekplan.md`,
  `commands/trekbrief.md`, `agents/research-orchestrator.md` (mode table,
  agent table, prompting block, summary line), `README.md` (feature prose, mode
  table, agent prose, the mermaid EXTERNAL node, the `* = conditional` legend,
  the optional-MCP requirement, the `--local` section), `CLAUDE.md`,
  `settings.json` (the whole `trekresearch.geminiBridge` block),
  `templates/research-brief-template.md` (the *Gemini Second Opinion* section
  and the sources-table `gemini` type), `templates/trekbrief-template.md`,
  `docs/architecture.md`, `docs/command-modes.md`.
 - **`/trekplan` high-effort Adversarial Pass 2 degrades explicitly.** It now
  emits a `## Adversarial Pass 2 (v5.1.1 high-effort)` section whose status
  reads **"unavailable, skipped"**, rather than failing or vanishing: a
  high-effort plan carrying no Pass 2 marker is indistinguishable from one whose
  Pass 2 crashed. `/trekresearch` high effort forces only `contrarian-researcher`
  always-on, which remains the complete adversarial coverage.
 - **Not removed:** `gemini_used` survives as a vestigial `/trekresearch` stats
  field pinned to `false`. Dropping it would break the observability export
  schema for existing consumers, and no consumer asked for that.
 - Agent inventory: **24 → 23 files** (20 spawnable + 3 orchestrator reference
  docs; `synthesis-agent` still dormant). Counts reconciled in `README.md` and
  `CLAUDE.md`; the `<example>`-block floor in
  `tests/lib/agent-frontmatter.test.mjs` moves 34 → 32 because an agent
  legitimately left the inventory — not because examples went missing from a
  surviving one.
 - Pinned by 5 new tests in `tests/lib/doc-consistency.test.mjs`, including a
  **known-positive control** asserting the historical records *still do* carry
  gemini references — so the empty result on live surfaces is a measurement and
  not a broken query. CHANGELOG entries and measurement/decision docs keep their
  references deliberately: they record what a past version did.
 ### Fixed
 - **`/trekreview` verdicts are fail-closed.** A finding removed by Pass 2 or
  Pass 3, and a reviewer whose payload was thrown away or never arrived, were
  arithmetically identical to a finding that never existed — all three pushed
  the verdict toward `ALLOW`. Measured before the fix: a BLOCKER with a
  101-character title → `ALLOW`; a payload carrying one ad-hoc `rule_key` was
  skipped whole at ingest, taking a valid BLOCKER sibling with it → `ALLOW`.
  `lib/review/coordinator-contract.mjs` now splits every removal two ways —
  **dropped** when the test refuted the finding as a claim about this codebase,
  **unverified** otherwise — and a non-empty `unverified` bucket, or a reviewer
  in `expectedReviewers` that did not report, forbids `ALLOW` (verdict `WARN`,
  with `allow_blocked_by` naming why). The rule never *raises* a verdict: the
  severity catalogue and the BLOCKER/MAJOR thresholds are untouched. New
  exports: `classifySuppression`, `REFUTING_REASONS`, `UNVERIFIED_REASONS`;
  `runContract` gains `unverified`, `missing_reviewers`,
  `unattributable_payloads` and `allow_blocked_by` (`suppressed` stays the
  union, so existing consumers keep their meaning). A payload that fails schema
  without carrying a `reviewer` name is counted as unattributable rather than
  reported as a reviewer called "unnamed reviewer" — naming one would invent an
  agent nobody launched and double-count with `expectedReviewers`.
  Mirrored in `agents/review-coordinator.md` (Pass 2/3 fate columns, the new
  §*Suppression is two-valued*, Pass 4 threshold table) and
  `commands/trekreview.md` (Phase 5 reviewer accounting → STOP; Phase 6).
  Driven test-first: 11 new tests in `tests/lib/coordinator-contract.test.mjs`,
  incl. a known-positive control proving `ALLOW` is still reachable.
 ### Docs
 - **T1 §6 PoC status corrected — it was already run and declined.**
  `docs/T1-cc26-delegated-orchestration.md` §8 item 3 still claimed both the §6
  synthesis-agent PoC and the §5 full bake-off were "designed but unbuilt". That
  was written in S7 and falsified the same afternoon by S12, which ran the §6
  PoC and recorded **Δ main-context (faithful flow) = 0.0%, NEGATIVE**
  (`docs/T1-synthesis-poc-results.md`; reproduce with
  `node scripts/synthesis-measure.mjs`). The stale wording caused the settled
  PoC to be re-ordered as new work on 2026-09-02, so it is struck rather than
  deleted and §6 gained an explicit **RUN AND DECLINED** status block. The
  finding is structural, not stochastic: `/trekplan` Phase 5 spawns the
  exploration swarm **foreground** (`commands/trekplan.md:158,338-341`), so the
  6–10 outputs are already resident in main before Phase 7 begins — delegating
  only the Phase-7 digest returns on top of them and evicts nothing. A saving
  needs the Phase-5 delivery moved off-main, which is §5.
 - **§5's environment premise updated (measured 2026-09-03, CC 2.1.259).** S12
  recorded the live bake-off as blocked for want of an `ANTHROPIC_API_KEY`;
  `claude -p --output-format stream-json` in fact runs on subscription auth with
  no API key and now emits a `subagent_stats` block (spawned / max_depth /
  refused) that did not exist in June — a ready-made classifier-interference
  counter for §5's feasibility guard. S12's other blocker (the installed plugin
  is the cache copy, so a repo-local agent is invisible to `claude -p`) stands
  unmeasured. This changes nothing about the §6 verdict; it lowers the cost of §5.
 - **`/trekresearch --engine deep-research`: document the real version window.**
  Claude Code **2.1.218** changed `/deep-research` to start only when the operator
  invokes it; from there the Skill tool refuses a model invocation outright with
  `disable-model-invocation` (measured in a real `--engine deep-research` run on
  2026-09-01 — the SC3 fallback to `swarm` held). The engine was built against
  2.1.196 and the prose still promised a path Claude Code has removed
  ("requires Claude Code 2.1.154+"). `commands/trekresearch.md` (flag bullet +
  pre-gate + fallback reason tokens), `docs/command-modes.md` and `README.md` now
  state the closed window `2.1.154 <= CC < 2.1.218` and name
  `disable-model-invocation` as the expected fallback reason on any current CC.
  The pre-gate gained an upper ceiling (still a numeric comparison, not a string
  one). **The flag is kept** as an additive opt-in that never hard-fails; no
  `lib/` change, no adapter-contract change, no default change, no version bump.
  Pinned by `tests/lib/doc-consistency.test.mjs`.
 ## v5.10.0 — 2026-08-18 — STORM bounded research loop (default-off) + the Agent-tool `name` spawn defect
 Additive. Every new research mechanism ships **inert**: `VOYAGE_STORM_ENABLED` is
--- a/CLAUDE.md
+++ b/CLAUDE.md
@ -65,8 +65,9 @@ Full flag reference for each command (modes, `--gates`, `--profile`, breaking ch
 | community-researcher | opus | Community experience: issues, blogs, discussions |
 | security-researcher | opus | CVEs, audit history, supply chain risks |
 | contrarian-researcher | opus | Counter-evidence, overlooked alternatives |
 | gemini-bridge | opus | Gemini Deep Research second opinion (conditional) |
-> **Inventory (S33 reconcile, updated v5.10.1).** 23 agent files = **20 spawnable** (one, `synthesis-agent`, ships **dormant** — Δ≈0, wired to nothing) **+ 3 orchestrator reference docs** (`planning-/research-/review-orchestrator` document the inline `/trek*` workflow, not spawnable capabilities). All 23 stay `model: opus` (operator pin `40d8742`); the glue/mechanical/retrieval/dormant roles were reconsidered for a sonnet downgrade and **kept opus** — decision record: `docs/voyage-vs-cc-balance-analysis.md` §10.
+> **Inventory (S33 reconcile).** 24 agent files = **21 spawnable** (one, `synthesis-agent`, ships **dormant** — Δ≈0, wired to nothing) **+ 3 orchestrator reference docs** (`planning-/research-/review-orchestrator` document the inline `/trek*` workflow, not spawnable capabilities). All 24 stay `model: opus` (operator pin `40d8742`); the glue/mechanical/retrieval/dormant roles were reconsidered for a sonnet downgrade and **kept opus** — decision record: `docs/voyage-vs-cc-balance-analysis.md` §10.
 > **Model & effort.** `opus` = Opus 4.8 (default reasoning effort `high`); `sonnet` = Sonnet 4.6; `fable` = Fable 5 (Mythos-class, above Opus — reasoning effort inherits from the session; xhigh requires a session-level setting). Select agents carry native per-spawn `effort:` (retrieval → `medium`, adversarial-reasoning → `high`) — a different axis from brief `phase_signals.effort` (orchestration shape: which agents/passes run). Per-agent table + axes → `docs/profiles.md` §Model & effort axes.
--- a/README.md
+++ b/README.md
@ -2,7 +2,7 @@
 Contract-driven Claude Code pipeline: brief, research, plan, execute, review. Agent swarms, research triangulation, adversarial review, multi-session resumption.
-![Version](https://img.shields.io/badge/version-5.10.1-blue)
+![Version](https://img.shields.io/badge/version-5.10.0-blue)
 ![License](https://img.shields.io/badge/license-MIT-green)
 ![Platform](https://img.shields.io/badge/platform-Claude%20Code-purple)
@ -195,7 +195,7 @@ If you say "stop" or "enough" during Phase 4, the current review findings are su
 Deep, multi-phase research that combines local codebase analysis with external knowledge. Uses specialized agent swarms to investigate multiple dimensions in parallel, then triangulates findings.
-A parallel swarm of up to 5 local + 4 external agents investigates 3–8 research dimensions. Findings are triangulated (local vs. external, confidence per dimension, contradictions flagged) and synthesized into a structured research brief.
+A parallel swarm of up to 5 local + 4 external agents investigates 3–8 research dimensions, with optional Gemini Deep Research as an independent second opinion. Findings are triangulated (local vs. external, confidence per dimension, contradictions flagged) and synthesized into a structured research brief.
 Output:
 - With `--project <dir>`: `{dir}/research/{NN}-{slug}.md` (auto-incremented index)
@ -205,14 +205,14 @@ Output:
 | Mode | Usage | Behavior |
 |------|-------|----------|
-| **Default** | `/trekresearch <question>` | Interview + research swarm (local + external), foreground |
+| **Default** | `/trekresearch <question>` | Interview + research swarm (local + external + Gemini), foreground |
 | **Project** | `/trekresearch --project <dir> <question>` | Write brief into `{dir}/research/NN-slug.md` |
 | **Quick** | `/trekresearch --quick <question>` | Interview (short) + inline research, no agent swarm |
-| **Local** | `/trekresearch --local <question>` | Only codebase analysis agents (skip external research) |
+| **Local** | `/trekresearch --local <question>` | Only codebase analysis agents (skip external + Gemini) |
 | **External** | `/trekresearch --external <question>` | Only external research agents (skip codebase analysis) |
 | **Foreground** | `/trekresearch --fg <question>` | No-op alias (foreground is default since v2.4.0) |
 | **Profile** | `/trekresearch --profile <name> <question>` | (v4.1.0) Pin model profile for the research phase. See [Profile system](#profile-system-v410). |
-| **Engine** | `/trekresearch --external --engine deep-research <question>` | Delegate the external phase to Claude Code's built-in `/deep-research` workflow. Only works on `2.1.154 <= CC < 2.1.218`; from **2.1.218** `/deep-research` is operator-invoked only (Skill tool: `disable-model-invocation`), so it always falls back to `swarm`. Never hard-fails. Default `swarm`. |
+| **Engine** | `/trekresearch --external --engine deep-research <question>` | Delegate the external phase to Claude Code's built-in `/deep-research` workflow; falls back to `swarm` if unavailable. Default `swarm`. |
 Flags combine: `--project <dir> --external`.
@ -234,7 +234,7 @@ how many turns it has used. Whether the loop becomes the default is decided by a
 pre-registered measurement, not by preference: see
 [`docs/storm-measurement.md`](docs/storm-measurement.md).
-Research uses up to 5 local agents (architecture-mapper, dependency-tracer, task-finder, git-historian, convention-scanner) and 4 external agents (docs-researcher, community-researcher, security-researcher, contrarian-researcher). Per-agent details in [`agents/`](agents/).
+Research uses up to 5 local agents (architecture-mapper, dependency-tracer, task-finder, git-historian, convention-scanner) and 4 external agents (docs-researcher, community-researcher, security-researcher, contrarian-researcher) plus the optional Gemini bridge for an independent second opinion. Per-agent details in [`agents/`](agents/).
 ---
@ -657,7 +657,7 @@ flowchart TB
    end
    subgraph RES["/trekresearch · Phase 4 — parallel"]
        RL["LOCAL: architecture-mapper · dependency-tracer<br/>task-finder · git-historian · convention-scanner*"]
-        RE["EXTERNAL: docs-researcher · community-researcher<br/>security-researcher* · contrarian-researcher*"]
+        RE["EXTERNAL: docs-researcher · community-researcher<br/>security-researcher* · contrarian-researcher* · gemini-bridge*"]
    end
    subgraph PL["/trekplan"]
        PLG["Phase 4b · brief-reviewer (gate)"]
@ -676,7 +676,7 @@ flowchart TB
    BR --> RES --> PL --> EX --> RV
 ```
-`* = conditional`: convention-scanner / test-strategist on medium+ codebases (50+ files); research-scout for unknown external tech; security-/contrarian-researcher when a leading recommendation forms (or always at `effort=high`); brief-conformance-reviewer skipped under `--quick`.
+`* = conditional`: convention-scanner / test-strategist on medium+ codebases (50+ files); research-scout for unknown external tech; security-/contrarian-researcher + gemini-bridge when a leading recommendation forms (or always at `effort=high`); brief-conformance-reviewer skipped under `--quick`.
 > **Which Claude Code primitive each phase uses — and the alternatives considered (Workflow substrate, delegated orchestrator, dormant synthesis-agent)** → see [docs/architecture.md §Primitives per step](docs/architecture.md#primitives-per-step-decision-matrix).
@ -837,6 +837,7 @@ For per-profile cost estimates, see [`docs/profiles.md`](docs/profiles.md).
 - [Claude Code](https://docs.anthropic.com/en/docs/claude-code) (CLI, desktop app, or web app)
 - Claude subscription with Opus access (Max plan recommended)
 - Optional: [Tavily MCP server](https://github.com/tavily-ai/tavily-mcp) for enhanced external research
 - Optional: a Gemini Deep Research MCP server exposing `gemini_deep_research`, `gemini_get_research_status`, `gemini_get_research_result`, `gemini_research_followup` for an independent second opinion (voyage calls these tools; any server implementing them works)
 ## Architecture
@ -844,8 +845,8 @@ Top-level layout:
 ```
 trekplan/
-├── agents/        23 specialized agents (all `model: opus`-pinned; per-phase model set by `--profile`)
+├── agents/        24 specialized agents (all `model: opus`-pinned; per-phase model set by `--profile`)
-│                  └ 20 spawnable (1 dormant: synthesis-agent, Δ≈0) + 3 orchestrator reference docs (not spawned)
+│                  └ 21 spawnable (1 dormant: synthesis-agent, Δ≈0) + 3 orchestrator reference docs (not spawned)
 ├── commands/      6 slash commands (trekbrief, trekresearch, trekplan, trekexecute, trekreview, trekcontinue) + trekendsession helper
 ├── templates/     Frontmatter templates for brief, research, plan, session, launch
 ├── hooks/         8 hooks (pre-bash, pre-write, pre-agent-cap, session-title, post-bash-stats, pre-compact-flush, post-compact-flush, otel-export)
@ -904,8 +905,8 @@ just run `--profile economy`, to put the whole pipeline on Sonnet.
 ### Disable external research
-`/trekresearch --local` skips Tavily and Microsoft Learn. To make
+`/trekresearch --local` skips Tavily, Microsoft Learn, and the
-`--local` the default, edit the front of
+Gemini bridge. To make `--local` the default, edit the front of
 `commands/trekresearch.md` Phase 1 and flip the default branch
 of the `--local` argument check. Or just always pass `--local` and
 document it in your team's CLAUDE.md.
--- a/SECURITY.md
+++ b/SECURITY.md
@ -2,14 +2,13 @@
 ## Reporting a vulnerability
-Report privately to <security@fromaitochitta.com> — do not open a public
+Open a **private** issue on Forgejo:
 issue. There is no SLA — this is a solo-maintained plugin — but
 acknowledged reports are usually triaged within 7 days.
-If you already have an account on the Forgejo instance, you may instead
+> https://git.fromaitochitta.com/open/ktg-plugin-marketplace
 open a **private** issue on the repo, tagged `security`:
-> https://git.fromaitochitta.com/open/voyage
+Tag it `security` and mark it private. Do not file public issues for
 unpatched vulnerabilities. There is no SLA — this is a solo-maintained
 plugin — but acknowledged reports are usually triaged within 7 days.
 ## Supported versions
--- a/agents/gemini-bridge.md
+++ b/agents/gemini-bridge.md
@ -0,0 +1,151 @@
 ---
 name: gemini-bridge
 description: |
  Use this agent when an independent second opinion from Gemini Deep Research is
  needed on a technology choice, architectural question, or complex research topic.
  Provides triangulation value by running a completely independent research path
  that can confirm or challenge findings from other agents.
 model: opus
 color: magenta
 tools: ["mcp__gemini-mcp__gemini_deep_research", "mcp__gemini-mcp__gemini_get_research_status", "mcp__gemini-mcp__gemini_get_research_result", "mcp__gemini-mcp__gemini_research_followup"]
 ---
 You are a bridge to Google Gemini Deep Research. Your role is to obtain an independent,
 thorough research result that provides triangulation value — a completely independent
 research path that can confirm or challenge findings from other agents.
 The value of this agent is INDEPENDENCE. Do not pre-bias Gemini with conclusions from
 other agents. Submit the research question cleanly so Gemini's findings stand on their
 own merits.
 ## Workflow
 ### 1. Check availability
 Attempt to call gemini_deep_research. If the tool is not available (MCP server not
 connected), return IMMEDIATELY with:
 ```
 ## Gemini Bridge Result
 **Status:** Unavailable
 **Reason:** Gemini MCP server not connected. Proceeding without second opinion.
 ```
 Do NOT error, block, or retry. Unavailability is an expected operational state.
 ### 2. Formulate query
 Take the research question and reformulate it for Gemini to maximize result quality:
 - Add context about what dimensions to cover (trade-offs, maturity, ecosystem, operational
  concerns, known failure modes, community consensus)
 - Use format_instructions to request structured output with clear sections, source citations,
  and explicit confidence levels per claim
 - Set parameters:
  - `research_mode`: "custom"
  - `source_tier`: 2
  - `research_window_days`: 90
 Example format_instructions to include:
 > "Structure your response with: Executive Summary, Key Findings (bullet points),
 > Trade-offs, Known Issues and Gotchas, Community Consensus, and Sources. For each
 > major claim, indicate your confidence level (high/medium/low) and cite the source."
 ### 3. Submit research
 Call `gemini_deep_research` with the reformulated query and parameters.
 ### 4. Poll for completion
 Call `gemini_get_research_status` repeatedly until the research completes:
 - Call the status tool, then call it again after it returns — repeat until done
 - Do not use bash or sleep commands — use repeated tool calls to simulate waiting
 - Continue polling until status is `"completed"` or `"failed"`
 - If `"failed"`: report the failure reason and return gracefully — do not retry
 - Timeout: if still running after 40 polls (~20 minutes of equivalent wait), report
  timeout and return whatever partial result is available
 ### 5. Retrieve result
 Call `gemini_get_research_result` with `include_citations: true`.
 ### 6. Optional follow-up
 If the result has clear gaps on specific dimensions that are directly relevant to the
 research question, call `gemini_research_followup` with a targeted follow-up question.
 Rules for follow-up:
 - Maximum 1 follow-up call
 - Only if there is a genuine gap — do not follow up out of habit
 - Make the follow-up question narrow and specific, not a re-statement of the original
 ### 7. Format output
 Structure the final result as:
 ```
 ## Gemini Bridge Result
 **Status:** Completed
 **Research duration:** {time taken}
 **Sources cited:** {count}
 ### Key Findings
 - {finding 1}
 - {finding 2}
 - {finding 3}
 ### Trade-offs and Known Issues
 - {trade-off or issue 1}
 - {trade-off or issue 2}
 ### Sources
 | # | Source | Relevance |
 |---|--------|-----------|
 | 1 | {URL}  | {one-line relevance} |
 ### Areas for Triangulation
 *Claims that should be cross-checked against local codebase analysis
 and other external agents:*
 - {claim 1 — check against local architecture}
 - {claim 2 — verify with community experience}
 - {claim 3 — validate against codebase constraints}
 ```
 ## Rules
 - **Never block the research pipeline.** If Gemini is slow or unavailable, return what
  you have with a clear status note.
 - **Do not interpret or editorialize.** Report Gemini's findings as-is, formatted for
  integration. Your job is formatting and delivery, not analysis.
 - **Flag "Areas for Triangulation"** — claims that the research-orchestrator or other
  agents should cross-check against local codebase analysis, team experience, or other
  external sources.
 - **Independence is the point.** Do not include findings from other agents in your query
  to Gemini. The value of a second opinion is that it is uninfluenced by the first.
 - **Cite everything.** Every major claim in the output must trace to a source in the
  Sources table. Remove claims that Gemini did not support with a source.
 - **Graceful degradation at every step.** Unavailable tool, failed research, timeout —
  all are handled with a clear status message and immediate return. Never leave the
  pipeline hanging.
 ## When to use — examples
 <example>
 Context: trekresearch launches gemini-bridge for an independent second opinion on a technology choice
 user: "/trekplan Should we use Kafka or NATS for our event streaming layer?"
 assistant: "Launching gemini-bridge for an independent second opinion on Kafka vs NATS."
 <commentary>
 Technology choice with significant architectural implications triggers gemini-bridge
 to provide an independent research path alongside local exploration agents.
 </commentary>
 </example>
 <example>
 Context: user wants deep research via Gemini on a complex architectural question
 user: "Get me a Gemini deep research on event sourcing patterns for distributed systems"
 assistant: "I'll use the gemini-bridge agent to run a deep research on event sourcing patterns."
 <commentary>
 Direct request for Gemini research on a complex architectural question triggers the agent.
 </commentary>
 </example>
--- a/agents/research-orchestrator.md
+++ b/agents/research-orchestrator.md
@ -61,16 +61,13 @@ Execute these phases in order. Do not skip phases.
 Based on the mode, determine which agent groups to launch:
-| Mode | Local agents | External agents |
+| Mode | Local agents | External agents | Gemini bridge |
-|------|-------------|-----------------|
+|------|-------------|-----------------|---------------|
-| `default` | Yes | Yes |
+| `default` | Yes | Yes | Yes (if enabled in settings) |
-| `local` | Yes | No |
+| `local` | Yes | No | No |
-| `external` | No | Yes |
+| `external` | No | Yes | Yes (if enabled) |
 | `quick` | N/A — handled inline by the command, not the orchestrator |
 > **No bridge agent (v5.10.1).** The third-party second-opinion bridge was dropped
 > by operator directive; there is no bridge column to select.
 **Local agents** (reuse existing plugin agents with research-focused prompts):
 | Agent | Purpose in research context |
@ -90,6 +87,12 @@ Based on the mode, determine which agent groups to launch:
 | `security-researcher` | CVEs, audit history, supply chain risks |
 | `contrarian-researcher` | Counter-evidence, overlooked alternatives, reasons to reconsider |
 **Bridge agent:**
 | Agent | Purpose |
 |-------|---------|
 | `gemini-bridge` | Independent second opinion via Gemini Deep Research |
 ### Phase 2 — Parallel research
 Launch ALL selected agents **in parallel** using the Agent tool — one message,
@ -124,6 +127,11 @@ Examples:
 Pass the research question, specific dimensions to investigate, and any context from
 the interview about what the user already knows or cares about.
 **Prompting gemini-bridge:**
 Pass the research question as-is. Do NOT pre-bias with findings from other agents —
 the value of Gemini is independence.
 ### Phase 3 — Targeted follow-ups
 Review all agent results. Identify knowledge gaps — areas where findings are thin,
@ -187,7 +195,7 @@ When done, your output message should contain:
 **Brief:** {brief path}
 **Confidence:** {overall confidence 0.0-1.0}
 **Dimensions:** {N} researched
-**Agents:** {N} local + {N} external
+**Agents:** {N} local + {N} external + {gemini status}
 ### Key Findings
 - {Finding 1}
@ -218,7 +226,7 @@ You can:
 - **Sources:** Every claim in the brief must cite a source (URL or file path).
  Never invent findings.
 - **Honesty:** If a question is trivially answerable, say so. Don't inflate research.
- **Graceful degradation:** If MCP tools are unavailable (Tavily, MS Learn), proceed
+- **Graceful degradation:** If MCP tools are unavailable (Tavily, Gemini), proceed
  with available tools and note the limitation in the brief metadata.
 - **Independence:** Do not pre-bias external agents with local findings or vice versa.
  The value is in independent perspectives that are THEN triangulated.
--- a/agents/review-coordinator.md
+++ b/agents/review-coordinator.md
@ -84,48 +84,31 @@ identical `(file, rule_key)` and `line == 0` collide.
 ### Pass 2 — HubSpot Judge filters (3 criteria)
-Remove findings that fail ANY of these filters. **The `Removed as` column
+Drop findings that fail ANY of these filters:
 is load-bearing** — see *Suppression is two-valued* below:
-| Filter | Test | Fails if | Removed as |
+| Filter | Test | Drop if |
-|--------|------|----------|------------|
+|--------|------|---------|
-| Succinctness | `title.length ≤ 100` and `detail.length ≤ 800` chars | Title is a paragraph or detail is a wall of text | `unverified` (`succinctness:title` / `succinctness:detail`) |
+| Succinctness | `title.length ≤ 100` and `detail.length ≤ 800` chars | Title is a paragraph or detail is a wall of text |
-| Accuracy | `file` resolves under the repo root AND `line` is plausible (≥ 0; ≤ file line count when known) | Path traversal escape, negative line, or impossibly large line number | **dropped** (`accuracy:refuted`) |
+| Accuracy | `file` resolves under the repo root AND `line` is plausible (≥ 0; ≤ file line count when known) | Path traversal escape, negative line, or impossibly large line number |
-| Actionability | `recommended_action` is non-empty AND begins with an imperative verb | Empty action, "consider …" hedges, or restating the title | `unverified` (`actionability:empty`) |
+| Actionability | `recommended_action` is non-empty AND begins with an imperative verb | Empty action, "consider …" hedges, or restating the title |
-Succinctness and Actionability read the finding's *packaging*; neither
+When dropping a finding, preserve a one-line note in the
 examines the claim, so neither can establish the finding is unreal. Accuracy
 does: a citation that escapes the repo root refutes the finding as a claim
 about this codebase.
 When removing a finding, preserve a one-line note in the
 `Suppressed Findings` body section so the user knows why the count
 shrank.
 ### Pass 3 — Cloudflare reasonableness (skipped in quick mode)
-Remove findings that fail ANY of these tests:
+Drop findings that fail ANY of these tests:
- **No file:line citation** → **dropped** (`no-citation`). `file` is empty,
+- **No file:line citation.** `file` is empty, or `line < 0`. Speculative
-  or `line < 0`. Speculative "code might break somewhere" findings name no
+  "code might break somewhere" findings have no anchor and are dropped.
-  location, so they make no checkable claim at all.
+- **Unknown rule_key.** `rule_key` is not in `RULE_CATALOGUE`. Reviewers
- **Unknown rule_key** → `unverified` (`unknown-rule_key`). `rule_key` is not
+  occasionally emit ad-hoc rule keys; the catalogue is the contract.
  in `RULE_CATALOGUE`. Reviewers occasionally emit ad-hoc rule keys; the
  catalogue is the contract, but a mislabelled finding is not a refuted one.
  *(High-effort mode does not reach this branch: Pass 3 is bypassed and the
  key is normalised to `PLAN_EXECUTE_DRIFT` and KEPT — see High-effort
  normalization below. The two fates never apply to the same input.)*
 - **Non-existent file.** `file` does not exist in the working tree AND
-  the diff does not show it as `(new file)`. Use Glob to verify. **This test
+  the diff does not show it as `(new file)`. Use Glob to verify.
  has three outcomes, not two:** Glob resolves and the file is absent from
  both tree and diff → **dropped** (`file-existence:refuted`); Glob resolves
  and the file is present → keep; **Glob cannot decide** (path outside the
  working tree, unreadable, or the tool errored) → `unverified`
  (`file-existence:indeterminate`). Never collapse *unresolvable* into
  *refuted*.
 - **Catalogue severity mismatch.** `severity` does not match the rule's
  catalogue tier (e.g., `MISSING_TEST` emitted as MINOR). Reset to the
-  catalogue tier; this is a correction, neither a drop nor an unverified.
+  catalogue tier; this is a correction, not a drop.
 In `quick` mode, skip this pass entirely. Note the skip in the
 Executive Summary so the reader knows reasonableness was not applied.
@ -142,32 +125,6 @@ purposes. This normalization happens BEFORE writing review.md,
 ensuring all `rule_key` values in the final review match the
 catalogue.
 ### Suppression is two-valued (fail-closed)
 Every removal in Pass 2 and Pass 3 carries one of two fates, and the
 distinction decides whether the review may come back clean:
 | Fate | Meaning | Weight in Pass 4 |
 |------|---------|------------------|
 | **dropped** | The test **refuted** the finding as a claim about this codebase. | None. It weighs nothing, correctly. |
 | **unverified** | The finding was removed **without** its claim ever being examined or settled. | Forbids `ALLOW`. |
 The rule is one sentence: **a removal is `dropped` only when the test
 refuted the finding; every other removal is `unverified`.** A reason you
 cannot place is `unverified` — the default fails closed.
 Why this exists: without it, a finding the coordinator could not
 substantiate is arithmetically identical to a finding that never existed,
 and both push the verdict toward `ALLOW`. The deterministic mirror of this
 rule, including the reason vocabulary, is
 `lib/review/coordinator-contract.mjs` (`classifySuppression`,
 `REFUTING_REASONS`, `UNVERIFIED_REASONS`) — prose and lib share one
 vocabulary on purpose.
 **Unverified findings are not counted into a severity tier.** Their severity
 is reviewer-asserted and was never substantiated; counting it would let an
 unexamined finding *raise* the verdict, which is invention.
 ### Pass 4 — Compute verdict
 Count findings by severity AFTER dedup and filtering. Verdict thresholds:
@ -176,19 +133,7 @@ Count findings by severity AFTER dedup and filtering. Verdict thresholds:
 |--------|---------|
 | `BLOCKER ≥ 1` | `BLOCK` |
 | `BLOCKER == 0` AND `MAJOR ≥ 1` | `WARN` |
-| `BLOCKER == 0` AND `MAJOR == 0` AND nothing `unverified` AND every reviewer reported | `ALLOW` |
+| `BLOCKER == 0` AND `MAJOR == 0` | `ALLOW` |
 | `BLOCKER == 0` AND `MAJOR == 0` AND (`unverified` non-empty OR a reviewer did not report) | `WARN` |
 The fail-closed row never RAISES a verdict — it only withholds the clean
 one. The worst case of a false `unverified` is `WARN` plus a stated reason;
 the worst case of the old behaviour was a silent `ALLOW` over a live
 BLOCKER.
 **When `ALLOW` is withheld, the Executive Summary's FIRST sentence must say
 so and name why** — e.g. "WARN: no blocking findings survived, but 1 finding
 could not be verified (succinctness:title) and brief-conformance-reviewer did
 not report." A withheld ALLOW that the reader cannot see is the same defect
 in a new place.
 Verdict is mechanical — never override. The verdict goes into the
 trailing JSON block AND the Executive Summary's first sentence.
@ -236,10 +181,8 @@ prefix). Flow-style `findings: [a, b]` breaks the frontmatter parser.
 5. `## Findings (MAJOR)` — one subsection per MAJOR finding.
 6. `## Findings (MINOR)` — one subsection per MINOR finding.
 7. `## Findings (SUGGESTION)` — one subsection per SUGGESTION finding.
-8. `## Suppressed Findings` (optional) — one line per finding removed by
+8. `## Suppressed Findings` (optional) — one-line per finding dropped by
-   Pass 2 or Pass 3, with the reason AND its fate, tagged `[dropped]` or
+   Pass 2 or Pass 3, with the reason.
   `[unverified]`. Unverified lines come first: they are the ones that
   withheld `ALLOW`.
 9. `## Remediation Summary` — bullet count per severity + 1 sentence on
   what /trekplan will consume.
@ -261,7 +204,6 @@ The LAST fenced block in the file is a `json` block:
 {
  "verdict": "BLOCK | WARN | ALLOW",
  "counts": { "BLOCKER": N, "MAJOR": N, "MINOR": N, "SUGGESTION": N },
  "allow_blocked_by": ["unverified:succinctness:title (1)", "missing-reviewer:brief-conformance-reviewer", "unattributable-payload (1)"],
  "findings": [
    {
      "id": "<40-char-hex>",
@ -301,14 +243,9 @@ for the ID list.
  the canonical 40-char SHA1 from `(file, line, rule_key, title)` using
  the algorithm in `lib/parsers/finding-id.mjs`. The frontmatter
  `findings:` list and the JSON block IDs must match.
- **Suppressed findings are accountable.** When you remove a finding via
+- **Suppressed findings are accountable.** When you drop a finding via
-  Pass 2 or Pass 3, log it in `## Suppressed Findings` with the reason and
+  Pass 2 or Pass 3, log it in `## Suppressed Findings` with the reason.
-  its fate (`[dropped]` / `[unverified]`). Silent drops break the audit
+  Silent drops break the audit trail.
  trail.
 - **Never spend an unexamined finding as evidence of a clean review.** If a
  removal did not refute the finding, it is `unverified` and `ALLOW` is off
  the table. This is the one place where you may not be minimal: when in
  doubt about a reason's fate, it is `unverified`.
 - **No invention.** Never add a finding that did not appear in the
  reviewer outputs. Never escalate a finding's severity beyond what the
  catalogue specifies.
--- a/commands/trekbrief.md
+++ b/commands/trekbrief.md
@ -386,8 +386,8 @@ The mapping table is canonical:
 - `high → {effort: high, model: opus}` (force opus for the high-confidence path)
 - `fable → {effort: high, model: fable}` (force Fable 5 for the max-quality path)
-The fable tier reuses `effort: high` semantics — full swarm, contrarian
+The fable tier reuses `effort: high` semantics — full swarm, contrarian +
-always-on; `EFFORT_LEVELS` is unchanged (Voyage effort is orchestration
+gemini always-on; `EFFORT_LEVELS` is unchanged (Voyage effort is orchestration
 shape, not model reasoning effort). Model reasoning effort is inherited from
 the session: Fable 5's default effort is `high`, NOT xhigh. To run xhigh, the
 operator sets it at session level via `/effort xhigh`, the `effortLevel`
@ -825,7 +825,7 @@ topic.
 > background orchestrator-agents could not spawn the research swarm —
 > before Claude Code 2.1.172 the harness did not expose the Agent tool to
 > sub-agents, so a background run silently degraded to single-context
-> reasoning without WebSearch / Tavily / WebFetch (see v2.4.0
+> reasoning without WebSearch / Tavily / WebFetch / Gemini (see v2.4.0
 > release notes). As of CC 2.1.172 sub-agents can spawn sub-agents (up to 5
 > levels deep), so a delegated redesign is under evaluation (see
 > `docs/cc-upgrade-2.1.181-decision-matrix.md`, W1/CC-26). Until then,
--- a/commands/trekexecute.md
+++ b/commands/trekexecute.md
@ -419,46 +419,19 @@ to stderr but do NOT block the stop; `progress.json` is still authoritative.
 `status: stopped`) so the next-session producer-mismatch check has both
 candidates available. Use the same combined ESM block pattern as Phase 8.
-### Check 2 — Plan file reaches every worktree
+### Check 2 — Plan file is tracked by git
-Worktrees are created from HEAD, so tracking the plan file is the cheapest way
+Run `git ls-files --error-unmatch {plan-path} 2>/dev/null`. If the plan file is
-to make it visible in each one. But the project directory may be **gitignored**
+untracked (exit code != 0):
 — `.claude/projects/` is tool-managed and local-only, and a repo that ignores it
 is normal, not exotic. `git add -f` is **not** the answer there: it would push
 operator-local artifacts into history, and into whatever remote the repo
 publishes to. When the plan file is ignored, Phase 2.6 Step 2a' (which copies
 brief/plan/research into each worktree) is the delivery path, and this check
 must step aside instead of failing.
 ```bash
-PLAN_PATH="{plan-path}"
+git add {plan-path}
-if git ls-files --error-unmatch "$PLAN_PATH" >/dev/null 2>&1; then
+git commit -m "chore: track plan file for parallel execution"
  PLAN_TRACKING="tracked"
 else
  git check-ignore -q "$PLAN_PATH"
  case "$?" in
    0) PLAN_TRACKING="ignored" ;;
    1) PLAN_TRACKING="untracked" ;;
    *) echo "Error: git check-ignore failed on $PLAN_PATH - a fatal probe is not an answer about ignore status." >&2
       exit 1 ;;
  esac
 fi
 if [ "$PLAN_TRACKING" = "untracked" ]; then
  git add "$PLAN_PATH"
  git commit -m "chore: track plan file for parallel execution"
 fi
 ```
-Report by outcome:
+Report: `Plan file committed for worktree visibility.`
-| `PLAN_TRACKING` | Report |
+This ensures every worktree created from HEAD will have the plan file.
 |---|---|
 | `tracked` | `Plan file already tracked.` |
 | `untracked` | `Plan file committed for worktree visibility.` |
 | `ignored` | `Plan file is gitignored — not forced into history. Step 2a' copies it into each worktree.` |
 Any other `git check-ignore` exit code is fatal and stops execution: a probe
 that failed is not a probe that answered "not ignored".
 ### Check 3 — Scope fence overlap validation
@ -505,7 +478,7 @@ If cleanup fails, report the manual commands and stop.
 After all 4 checks pass:
 ```
-Pre-flight: PASS (clean tree, plan reaches worktrees, no overlaps, no stale worktrees)
+Pre-flight: PASS (clean tree, plan tracked, no overlaps, no stale worktrees)
 ```
 ## Phase 2.6 — Multi-session orchestration (worktree-isolated)
@ -631,45 +604,27 @@ Insert this block AFTER the worktree-creation loop and BEFORE wave dispatch
 ```bash
 PROJECT_SOURCE="$(realpath "${PROJECT_DIR}")"
 REPO_ROOT_REAL="$(realpath "${REPO_ROOT}")"
 # Compute destination relpath: PROJECT_DIR relative to REPO_ROOT.
 # This makes $wt/$PROJECT_REL valid regardless of whether the operator
 # passed --project as relative (.claude/projects/...) or absolute.
-# python3 + os.path.relpath is stdlib and portable - see the note below the
+PROJECT_REL="$(realpath --relative-to="$REPO_ROOT" "$PROJECT_SOURCE")"
 # block for why no realpath flag may be used here.
 PROJECT_REL="$(python3 -c 'import os.path,sys; print(os.path.relpath(sys.argv[1], sys.argv[2]))' "$PROJECT_SOURCE" "$REPO_ROOT_REAL")"
 case "$PROJECT_REL" in
  ""|..*)
    echo "Error: cannot derive a project relpath inside the repo ($PROJECT_SOURCE vs $REPO_ROOT_REAL)." >&2
    exit 1 ;;
 esac
 for wt in "$WORKTREE_DIR"/session-*; do
  [ -d "$wt" ] || continue
  mkdir -p "$wt/$PROJECT_REL"
  cp "$PROJECT_SOURCE"/brief.md "$wt/$PROJECT_REL/"
  cp "$PROJECT_SOURCE"/plan.md  "$wt/$PROJECT_REL/"
-  if [ -d "$PROJECT_SOURCE/research" ]; then
+  [ -d "$PROJECT_SOURCE/research" ] && \
    cp -r "$PROJECT_SOURCE/research" "$wt/$PROJECT_REL/"
  fi
 done
 ```
-Do not "improve" the relpath line into `realpath --relative-to=...`. That flag
+Note: `realpath --relative-to` is GNU coreutils. macOS users without
-is GNU coreutils only; BSD `realpath` (the macOS default) rejects it, and the
+`coreutils` (Homebrew `brew install coreutils` provides `grealpath`) may
-failure is **silent** — the command substitution leaves `PROJECT_REL` empty, so
+substitute a Python fallback:
-`mkdir -p "$wt/"` and `cp ... "$wt//"` both succeed and drop `brief.md`/`plan.md`
+`python3 -c "import os.path,sys; print(os.path.relpath(sys.argv[1], sys.argv[2]))" "$PROJECT_SOURCE" "$REPO_ROOT"`.
 at the worktree root, where no child session looks for them. Measured on an
 Intel Mac 2026-08-31: `realpath --relative-to=... ` → `realpath: illegal option
 -- -`, while bare `realpath <path>` works; a whole wave ran with zero steps
 executed. Both `realpath` calls above are bare path resolution, which BSD and
 GNU handle identically; resolving both operands before the relpath is what keeps
 it correct when one side goes through a symlink (macOS `/var` → `/private/var`).
-Failure modes: an underivable relpath (empty, or outside the repo) aborts
+Failure mode: any `cp` failure exits the wave non-zero; reported via Step 4
-before anything is copied — better a loud stop than files delivered where no
+cleanup. Source: brief Constraint 2.
 session reads them; any `cp` failure exits the wave non-zero, reported via
 Step 4 cleanup. A project without `research/` is not a failure.
 Source: brief Constraint 2.
 **2b. Launch sessions in this wave (each in its own worktree):**
--- a/commands/trekplan.md
+++ b/commands/trekplan.md
@ -898,33 +898,22 @@ validator-only; this surface just makes the friendly hint readable.
 ### High-effort behavior (v5.1.1)
-When `phase_signal_result.effort == 'high'` for the `plan` phase, the
+When `phase_signal_result.effort == 'high'` for the `plan` phase, after
-additional adversarial Pass 2 on the post-revision plan is
+Phase 9 (plan-critic + scope-guardian dedup pass) runs to completion on
-**unavailable, skipped**. Append the section header to plan.md before
+the post-revision plan, run an ADDITIONAL `gemini-bridge` plan-review
-the trailing JSON block with exactly that status, so a reader can tell
+pass on the post-revision plan. Surface its findings as a separate
-a deliberately dropped pass from one that failed:
+`## Adversarial Pass 2 (gemini-bridge, v5.1.1 high-effort)` section
 appended to plan.md before the trailing JSON block.
-```
+Rationale (per risk-assessor finding + Decision B substitution
-## Adversarial Pass 2 (v5.1.1 high-effort)
+2026-05-14): the originally-considered "extra plan-critic-iterasjon"
 risked a revision-loop because plan-critic dedup keys on
 `(file, line, rule_key)` triplets and post-revision line numbers shift.
 The gemini-bridge pass is independent (different agent, different
 perspective) and does not re-tread the same dedup space — it surfaces
 genuinely new findings rather than re-emitting closed ones.
-**Status:** unavailable, skipped — the third-party bridge agent that
+Standard and low effort: do NOT run the additional pass.
 ran this pass was dropped in v5.10.1 by operator directive. No
 substitute is wired. Phase 9 (plan-critic + scope-guardian) is the
 complete adversarial coverage for this plan.
 ```
 Do NOT fail, retry, or silently omit the section: a high-effort plan
 that carries no Pass 2 marker is indistinguishable from one whose Pass
 2 crashed.
 History (why the slot exists): Decision B, 2026-05-14, substituted a
 bridge pass for an "extra plan-critic-iterasjon" because plan-critic
 dedup keys on `(file, line, rule_key)` triplets and post-revision line
 numbers shift, risking a revision-loop. Dropping the bridge does not
 revive that option — re-running plan-critic carries the same loop risk
 it always did.
 Standard and low effort: do NOT emit the section at all.
 ## Hard rules
--- a/commands/trekresearch.md
+++ b/commands/trekresearch.md
@ -32,7 +32,7 @@ Supported flags:
 1. `--quick` — lightweight research, no agent swarm. The command itself does
   3-5 targeted searches inline. Set **mode = quick**.
-2. `--local` — only codebase research. Skip external research agents.
+2. `--local` — only codebase research. Skip external agents and gemini bridge.
   Set **scope = local**.
 3. `--external` — only external research. Skip codebase analysis agents.
@ -87,14 +87,9 @@ Supported flags:
   behavior). `swarm` runs Voyage's own external-research agent swarm;
   `deep-research` delegates the external phase to Claude Code's built-in
   `/deep-research` dynamic workflow and adapts its report into the research-brief
-   schema. **The delegation only works inside a closed version window:
+   schema (requires Claude Code 2.1.154+ and dynamic workflows enabled; falls back
-   `2.1.154 <= CC < 2.1.218`** (dynamic workflows enabled). From Claude Code
+   to `swarm` and notes the fallback if unavailable — never hard-fails). Orthogonal
-   **2.1.218** `/deep-research` starts only when the operator invokes it, and the
+   to `--profile`/`phase_signals`; only affects the external phase. Set
   Skill tool refuses a model invocation outright (`disable-model-invocation`), so
   on 2.1.218 or newer the engine **always** falls back to `swarm`. The flag is
   kept as an additive opt-in and **never hard-fails**: outside the window it
   degrades to `swarm` and logs the reason. Orthogonal to
   `--profile`/`phase_signals`; only affects the external phase. Set
   **engine = {swarm|deep-research}** (the *requested* engine).
 Flags can be combined:
@ -121,7 +116,7 @@ Usage: /trekresearch <research question>
 Modes:
  default       Interview → foreground research (local + external) → brief
  --quick       Interview (short) → inline research (no agent swarm)
-  --local       Only codebase analysis agents (skip external research)
+  --local       Only codebase analysis agents (skip external + Gemini)
  --external    Only external research agents (skip codebase analysis)
  --fg          No-op alias (foreground is the only mode as of v2.4.0)
  --project     Write brief into an trekbrief project folder (auto-indexed)
@ -240,8 +235,8 @@ Then continue to the next phase inline.
 > did not expose the Agent tool to sub-agents, so an orchestrator launched
 > with `run_in_background: true` could not spawn the documented research
 > swarm (`docs-researcher`, `community-researcher`, etc.) and silently
-> degraded to single-context reasoning without WebSearch / Tavily / WebFetch.
+> degraded to single-context reasoning without WebSearch / Tavily / WebFetch
-> As of CC 2.1.172 sub-agents can spawn sub-agents (up to 5 levels
+> / Gemini. As of CC 2.1.172 sub-agents can spawn sub-agents (up to 5 levels
 > deep), so that block no longer holds — a delegated redesign is under
 > evaluation (see `docs/cc-upgrade-2.1.181-decision-matrix.md`, W1/CC-26).
 > Until then, running the phases inline in main context keeps the swarm
@ -333,13 +328,9 @@ changes nothing (SC1). Keep the native-swarm anchors intact ("in parallel",
 1. **Coarse pre-gate (best-effort, NOT a trust signal).** `Bash: claude --version`;
   parse the leading `X.Y.Z` (e.g. from `2.1.196 (Claude Code)`) and compare
-   numerically against **both ends** of the supported window
+   numerically against `2.1.154` — split each on `.` and compare major, then minor,
   `2.1.154 <= version < 2.1.218` — split each on `.` and compare major, then minor,
   then patch as integers (do NOT string-compare; lexical comparison mis-orders
-   multi-digit patch numbers). If the version is `< 2.1.154` (below the
+   multi-digit patch numbers). If the version is `< 2.1.154`, OR if
   dynamic-workflows floor), OR `>= 2.1.218` (the ceiling: `/deep-research` starts
   only when the operator invokes it, and the Skill tool refuses a model invocation
   with `disable-model-invocation` — see step 4's reason token), OR if
   `disableWorkflows: true` / `CLAUDE_CODE_DISABLE_WORKFLOWS=1` is set, skip to the
   fallback (step 4). **If `claude` is not on PATH inside the Bash tool (possible
   under `claude -p`) or the version cannot be parsed, treat the pre-gate as
@ -356,18 +347,14 @@ changes nothing (SC1). Keep the native-swarm anchors intact ("in parallel",
   `/deep-research` report actually landed in context — substantive findings with
   citations, not an empty/denied/errored turn and not bare error text. This check
   must be **robust to all failure manifestations** (workflow disabled, approval
-   denied, runtime error, empty output, or the Skill tool refusing with
+   denied, runtime error, empty output), because the disabled-headless behavior is
-   `disable-model-invocation` on CC 2.1.218+), because the disabled-headless
+   undocumented: no recognizable cited report in context → fall back, regardless of
-   behavior is undocumented: no recognizable cited report in context → fall back,
+   how the failure surfaces.
   regardless of how the failure surfaces.
 4. **On no real report (fallback):** set `effective_engine = swarm`, run the swarm
   blocks below, and **log the fallback at this decision point** — print
   `Engine: deep-research → swarm (fallback: <reason>)` and carry the reason into the
-   Phase-8 Present summary and the brief's `## Executive Summary`. Known reason
+   Phase-8 Present summary and the brief's `## Executive Summary`. **NEVER fabricate
   tokens: `disable-model-invocation` (CC >= 2.1.218 — the expected reason on any
   current Claude Code), `version-below-floor`, `workflows-disabled`,
   `no-cited-report`. **NEVER fabricate
   or synthesize a substitute report** — a structurally-valid-but-invented brief
   passes the structure-only validator and silently poisons `/trekplan`; that is the
   worst outcome of this feature.
@ -391,6 +378,11 @@ Launch the new research-specialized agents:
 For each external agent, pass: the research question, specific dimensions to
 investigate, and any context from the interview.
 ### Bridge agent (scope = both or external, if enabled)
 Launch `gemini-bridge` with the research question. Do NOT include findings from
 other agents — the value of Gemini is independence.
 ### Launch rules
 - Launch ALL selected agents **in parallel** in a single message
@ -751,7 +743,7 @@ Present a summary to the user:
 **Engine (effective):** {swarm | deep-research}{, with fallback reason if it fell back}
 **Confidence:** {overall confidence 0.0-1.0}
 **Dimensions:** {N} researched
-**Agents:** {N} local + {N} external
+**Agents:** {N} local + {N} external + {gemini: used | unavailable | skipped}
 ### Key Findings
 - {Finding 1}
@ -796,7 +788,7 @@ Record format (one JSON line):
  "unique_sources": {N},
  "agents_local": {N},
  "agents_external": {N},
-  "gemini_used": false,
+  "gemini_used": {true|false},
  "confidence": {0.0-1.0},
  "contradictions": {N},
  "open_questions": {N}
@ -887,16 +879,15 @@ commands surface, don't re-enforce.
 When `phase_signal_result.effort == 'high'` for the `research` phase,
 run the FULL swarm regardless of normal triggering rules: 5 local
-agents + 4 external agents, AND force `contrarian-researcher` to
+agents + 4 external agents + 1 bridge agent, AND force
-always-on. Normally `contrarian-researcher` triggers conditionally when
+`contrarian-researcher` AND `gemini-bridge` to always-on. Normally
-a leading recommendation is emerging from initial agents; in
+`contrarian-researcher` triggers conditionally when a leading
-high-effort mode it runs unconditionally so the final brief always
+recommendation is emerging from initial agents; in high-effort mode it
-carries an adversarial counter-evidence pass.
+runs unconditionally so the final brief always carries an adversarial
-
+counter-evidence pass. Similarly, `gemini-bridge` normally activates on
-The former third-party second-opinion bridge pass is **unavailable,
+significant architectural questions or when triangulation value is
-skipped** (dropped in v5.10.1 by operator directive). High effort no
+high; in high-effort mode it runs unconditionally to provide an
-longer has a bridge arm; `gemini_used` is a vestigial stats field
+independent second opinion.
 pinned to `false`.
 High effort additionally arms the Phase 5 bounded follow-up loop (max 3
 turns per under-illuminated dimension, budgeted by
@ -943,7 +934,7 @@ Low effort: inline research only, no agent swarm (existing
  local finding from steering an external agent's question. Attributing the bias
  risk to it left that risk with no control while the text read as though it had
  one.
- **Graceful degradation:** If MCP tools are unavailable (Tavily, MS Learn),
+- **Graceful degradation:** If MCP tools are unavailable (Tavily, Gemini, MS Learn),
  proceed with available tools and note limitations in brief metadata.
 - **Cost:** Model resolution at Agent-spawn sites is a three-layer fallback:
  brief `phase_signals[<phase>].model` > `profile.phase_models[<phase>]` >
--- a/commands/trekreview.md
+++ b/commands/trekreview.md
@ -249,37 +249,6 @@ do not feed unvalidated findings to the coordinator.
 In `quick` mode, launch only `code-correctness-reviewer`. The Executive
 Summary will note the brief-conformance pass was skipped.
 ### Reviewer accounting — every expected reviewer MUST report
 Write down the expected reviewer set BEFORE the spawn: both reviewers in
 default mode, `code-correctness-reviewer` alone in `quick` mode. After the
 spawn, account for each one by name.
 **Zero findings from a silent reviewer is indistinguishable from zero findings
 from a clean diff** — unless you check. A reviewer is *accounted for* only when
 it returned a payload that validated. Three ways it fails to:
 | Failure | Handling |
 |---------|----------|
 | Output fails the schema after the 2 bounded re-asks | STOP (already specified above) |
 | Returned no final message at all | Re-ask that reviewer **once**. Still nothing → STOP. |
 | Was never launched (spawn error, wrong mode) | STOP. |
 **On STOP: name the reviewer and the failure, and do not proceed to Phase 6.**
 Do not let the coordinator compute a verdict over a review one of whose
 reviewers never spoke — the count would be complete-looking and wrong. This is
 the same shape as the schema branch above ("do not feed unvalidated findings to
 the coordinator"), applied to the other two ways a reviewer can go missing.
 A reviewer that ran but never delivered is most often the return-channel
 defect: check `~/.claude/projects/<proj>/<session>/subagents/agent-*.jsonl` for
 its final assistant block before re-asking, and confirm no `name` parameter was
 passed at the spawn (see the warning at the top of this phase).
 If you proceed anyway under an explicit operator instruction, pass the expected
 set to the coordinator as `expectedReviewers` so the missing reviewer at least
 forbids `ALLOW` (`lib/review/coordinator-contract.mjs`, `missing_reviewers`).
 ## Phase 6 — Coordinator dedup + verdict
 Launch `review-coordinator` (Agent tool) with the merged findings array
@ -290,20 +259,10 @@ The coordinator runs the 4-pass process documented in
 1. **Dedup** by `(file, line, rule_key)` triplet.
 2. **HubSpot Judge filters** — Succinctness, Accuracy, Actionability.
-3. **Cloudflare reasonableness** — remove speculative or catalogue-violating
+3. **Cloudflare reasonableness** — drop speculative or catalogue-violating
   findings (skipped in `quick` mode).
 4. **Verdict** — BLOCK / WARN / ALLOW per the threshold table.
 **Fail-closed.** Every removal in Pass 2 and Pass 3 is either
 **dropped** (the test refuted the finding as a claim about this codebase) or
 **unverified** (the finding was removed without its claim ever being settled).
 A non-empty `unverified` bucket forbids `ALLOW`; the verdict becomes `WARN` and
 the Executive Summary's first sentence must say why. The fail-closed rule never
 raises a verdict — it only withholds the clean one. Fate table, reason
 vocabulary, and the `allow_blocked_by` field: `agents/review-coordinator.md`
 §*Suppression is two-valued*, mirrored deterministically in
 `lib/review/coordinator-contract.mjs`.
 The coordinator's output is the full review.md content — frontmatter +
 body sections + trailing JSON block. Do NOT re-run the reviewers based
 on the coordinator's output.
--- a/docs/BRIEF-vurdering-v2.md
+++ b/docs/BRIEF-vurdering-v2.md
@ -1,295 +0,0 @@
 # BRIEF — external assessment (VURDERING-V2 §5.12), verified against this repo
 **Status:** assessment complete. **No implementation in this session** (scope-guard).
 **Source under review:** `claude-playlist-corpus/docs/VURDERING-V2.md` §5.12 (+ §3 G1–G8, §4),
 transferred via that repo's `docs/OVERFORING-V2.md` §5. It is an external recommendation
 built on 442 video analyses; its repo facts come from a subagent survey (2026-07-17) that
 is **not** re-verified and in which at least one error was already demonstrated.
 **Method.** Every asserted GAP is marked **BEKREFTET** (confirmed) / **AVKREFTET**
 (refuted) / **ENDRET** (true in altered form) only after checking it against code in this
 repo. Every Claude Code *feature* claim is checked against the OKF bundle at
 `claude-code-llm-wiki/bundle/`, index-first — **where V2 and the bundle disagree, the
 bundle wins**. Absences are positive-controlled before being reported (a grep that finds
 nothing is a measurement, not a fact).
 ---
 ## TL;DR
 Of six recommended tiltak, **one is a real defect** (fail-open verdict computation),
 **one is unmeasured rather than unresolved** (the Workflow port), **three are already
 decided or already partly built**, and **one is a cheap confirmed gap**.
 The single most consequential correction: V2 presents the Workflow port as the answer to
 this repo's "main-context relief is unmeasured" caveat. The bundle confirms the
 *mechanism*, but **this repo has still never measured main-context relief on a large
 fan-out** — the one bake-off that exists measured *subagent* tokens and says so. The port
 is therefore neither refuted nor justified. The next step is a measurement, not a port,
 and the measurement design already exists in `docs/T1-cc26-delegated-orchestration.md`.
 ---
 ## Tiltak 1 — Port orchestration to dynamic Workflows
 **Verdict: split per command. Do not issue one global verdict.**
 ### `/trekreview` — **ENDRET** (already ported, and measured)
 The port exists: opt-in `--workflow`, `scripts/trekreview-armB.workflow.mjs`, documented at
 `commands/trekreview.md:270` (*§ Phase 5–6 via the Workflow substrate*). The S10 bake-off
 (`docs/T2-bakeoff-results.md`, ≥3 runs/arm) found it **fidelity-equivalent**: verdict-match
 1.0, issue-coverage 100%, `(file,rule_key)` jaccard ≥ within-arm. It stays opt-in because
 the Workflow tool raises the consumer floor to Claude Code 2.1.154+, and the prose path
 keeps the lower floor.
 ### `/trekplan`, `/trekresearch` — **BEKREFTET as unported, UNMEASURED as beneficial**
 Both still orchestrate their swarms from main context (`commands/trekplan.md:406`,
 `commands/trekresearch.md:388`). That part of V2 is correct.
 What is **not** established is V2's stated effect — that Workflow answers the
 main-context-relief caveat. Two measurements exist in this repo and **neither measures it**:
 | Measurement | What it actually measured | Bearing on the relief claim |
 |---|---|---|
 | `docs/T2-bakeoff-results.md:154` | **Subagent** tokens: Arm A median 92.8k vs Arm B 96.9k ⇒ **+4.4%** | None. The same line states Arm A's main-context hand-orchestration cost is **uncounted**. |
 | `docs/T1-synthesis-poc-results.md` | Δ **main-context** ≈ **0%** — for delegating *Phase 7 synthesis alone* | Negative, but out of scope: that doc itself names a **Phase-5 redesign** as the prerequisite for any relief. |
 So the Phase-5 swarm — the case V2 is actually recommending — has never been measured in
 either arm, in either command.
 **The measurement already has a design.** `docs/T1-cc26-delegated-orchestration.md:118`
 defines the gate metric (Δ main-context tokens resident in the main session at
 plan-complete, from stream-json `usage`) with pre-registered bars: adopt at ≥30%, reject
 below 15%. Running Arm A vs Arm B once against the trekplan swarm settles this.
 **Cost framing must be corrected.** The bundle states plainly that a workflow run "can use
 meaningfully more tokens than working through the same task in conversation"
 (`bundle/concepts/docs-en-workflows.md:364`) — consistent with the measured +4.4%. The port
 cannot be justified as a cost saving. **Main-context headroom is the only defensible
 motive, and it is exactly the unmeasured quantity.**
 **Recommendation:** measure before porting. Treat V2's tiltak 1 as a *hypothesis with a
 ready-made experiment*, not as an approved change.
 ---
 ## Tiltak 2 — Codify the holdout rules
 **Verdict: (a) and (b) satisfied in practice but uncodified · (c) BEKREFTET, a real defect ·
 (d) ENDRET.** This is the tiltak with genuine substance.
 ### (a) Reviewers must never be forks — **AVKREFTET as a live defect, BEKREFTET as uncodified**
 Reviewers are spawned as plain subagents (`voyage:brief-conformance-reviewer`,
 `voyage:code-correctness-reviewer`). `subagent_type: "fork"` appears nowhere in `commands/`
 or `agents/`; the only `subagent_type` in the tree is `"Explore"` at
 `commands/trekplan.md:501`, an exploration spawn, not a review spawn. The bundle confirms
 the rule is real — a forked subagent "inherits your full conversation context instead of
 starting fresh" (`bundle/concepts/docs-en-agents.md:37`). Voyage satisfies it by habit;
 nothing written forbids a future edit from breaking it.
 ### (b) Explicit "you shall not see" list — **AVKREFTET as a live defect, BEKREFTET as uncodified**
 Phase 5 feeds reviewers exactly four things: the Phase-3 unified diff, the triage map, the
 brief path, and the rule catalogue (`commands/trekreview.md`, Phase 5 input list).
 Implementation plans and commit messages are **not** fed; `git log` appears only to compute
 the diff range and to print a suggested narrowing command (`commands/trekreview.md:132`,
 `:181`). Again: correct today, unprotected tomorrow.
 ### (c) Fail-closed verdicts — **BEKREFTET. This is the one real defect.**
 `agents/review-coordinator.md` computes the verdict mechanically from the findings that
 *survive* filtering:
 - **Pass 2** (HubSpot judge) and **Pass 3** (Cloudflare reasonableness) **drop** findings
  that fail their tests (`agents/review-coordinator.md:87`, `:101`).
 - **Pass 4** counts the survivors: `BLOCKER ≥ 1 → BLOCK`; else `MAJOR ≥ 1 → WARN`; else
  `ALLOW` (`:132`). "Verdict is mechanical — never override."
 There is no third state. Three consequences follow:
 1. **A finding that cannot be substantiated is dropped, and a dropped finding contributes
   zero — which moves the verdict toward ALLOW.** Pass 3's non-existent-file test ("`file`
   does not exist in the working tree AND the diff does not show it as `(new file)`") has
   no inconclusive branch: unresolvable and refuted are treated identically. This is
   structurally the failure V2 attributes to its source case, where "not E2E testable"
   passed as approved.
 2. **A reviewer that returns nothing produces the same result.** Phase 5 has no
   empty-return handling — grep for `empty|no findings|did not return|abort` over
   `commands/trekreview.md` returns 0 (positive control: `Phase 5` = 7 hits, `reviewer` =
   24 hits, so the file and the query are both live). Zero findings from a silent reviewer
   is indistinguishable from zero findings from a clean diff. This is the same surface as
   the idle-agent hole recorded in STATE's open decisions.
 3. **One fail-closed branch already exists, and it is the pattern to copy.** When reviewer
   output fails schema validation, Phase 5 allows 2 bounded re-asks and then *stops*: "do
   not feed unvalidated findings to the coordinator." That is the correct shape, applied to
   one failure mode only.
 **The defect is codified, not merely prose — which makes it testable.** The four passes
 exist deterministically in `lib/review/coordinator-contract.mjs`, and
 `computeVerdict(findings)` (`:184`–`:193`) counts *only* the findings handed to it.
 `runContract` hands it `reasoned.kept` (`:206`); `suppressed` and `skipped` are returned in
 the result object but **carry no weight in the verdict**. So a dropped finding is
 arithmetically identical to a finding that never existed.
 **Minimal fix shape (not implemented here):** an `unverified` bucket that is neither kept
 nor dropped, plus one rule — *a non-empty `unverified` bucket forbids ALLOW* — and a
 Phase-5 check that a reviewer actually returned. Both are additive; neither touches the
 severity catalogue or the existing thresholds. Because the logic is deterministic ("No LLM,
 no network, no time, no randomness", `:23`) and already has
 `tests/lib/coordinator-contract.test.mjs`, this can be driven test-first under the Iron
 Law — a failing test asserting that an unverifiable BLOCKER-severity finding cannot yield
 ALLOW is writable before any production change.
 ### (d) Presume-failure framing — **ENDRET**
 Both reviewers are already adversarially framed: "Adversarial reviewer" in each
 `description`, "You never praise", "You never say 'looks good'"
 (`agents/code-correctness-reviewer.md:36`, `agents/brief-conformance-reviewer.md:35`). What
 is absent is specifically the presume-failure formulation V2 names ("this agent was lazy —
 find out why"). This is a wording change with no measurement behind it in the corpus; treat
 it as optional polish, not a gap.
 ---
 ## Tiltak 3 — Lightweight lane — **ENDRET** (narrower gap than stated, already tracked)
 A lightweight path exists: `--quick` on both `/trekbrief` and `/trekplan`, and
 `/trekplan --quick` already skips the Phase-5 exploration swarm — "Skip agent swarm; use
 lightweight Glob/Grep scan and go directly to planning + adversarial review"
 (`commands/trekplan.md:122`). The same code-path is reachable a second way: a brief
 carrying `effort == 'low'` activates it without the flag (`:885`). What it does **not** do
 is bypass the 2.2 ceremony —
 `commands/trekbrief.md:126` states the framing question "is asked even in `--quick` mode",
 and a 2.2 brief still requires `framing` and a `## TL;DR` (`:472`, `:474`).
 So the accurate gap is not "no lightweight path" but "the lightweight path still pays the
 2.2 toll". Already carried as open operator decision 1 in STATE — this is not a discovery.
 ---
 ## Tiltak 4 — Definition-of-done as an object — **BEKREFTET**
 `commands/trekbrief.md` contains no notion of evidence, proof, verifier, residual risk, or
 next-step owner (grep for `evidence|proof|screenshot|verifier|residual risk|owner` → 0;
 positive control: `Success Criteri` = 5 hits). The gap is real.
 **Cost note V2 does not carry:** the brief schema is **Handover 1**, a public contract
 (`docs/HANDOVER-CONTRACTS.md`). Adding required DoD fields is a breaking change for
 downstream consumers, i.e. a `brief_version` bump with a gate, not an edit. That moves it
 out of "cheap win" and into planned work.
 ---
 ## Tiltak 5 — Prune, don't build — **mostly ALREADY DECIDED**
 - **`synthesis-agent`**: already dormant and labelled as such (`agents/synthesis-agent.md:3`),
  with the measurement that justified it (`docs/T1-synthesis-poc-results.md`, Δ≈0,
  DECLINED). Retiring versus keeping was decided in favour of keeping it as a
  re-measurable building block. V2 recommends a decision that has been made and recorded.
 - **24 all-opus agents**: 24 of 24 agent files carry `model: opus` (verified by count).
  This is an **operator pin**, not drift — commit `40d8742` "pin all sub-agents to Opus
  permanently (operator request)" — and the reconsideration V2 asks for already happened
  and is written down in `docs/voyage-vs-cc-balance-analysis.md` §10. Re-opening it is an
  operator decision, not an analysis task.
 - **`gemini-bridge`**: **BEKREFTET**, and the cheapest real item in this tiltak. It is
  already flagged `THIN_WRAP` / `DROP→NATIVE` in the same analysis (§V09, lines 81, 121,
  149) *and* it is broken at the engine (the `gemini-mcp` server fails deterministically on
  an SDK/API mismatch). Fix-or-drop is decidable today.
 ---
 ## Tiltak 6 — LSP as a harness component — **BEKREFTET**
 `LSP` returns 0 hits across `commands/`, `agents/`, `lib/`, `docs/`. Voyage navigates with
 Glob/Grep, as V2 says. One qualifier: LSP is a harness-level capability available to
 agents, not a voyage feature — adoption is mostly permitting and prompting agents to use
 it, not code in this repo. Value is highest in large target repos, which is where voyage's
 exploration swarm actually runs.
 ---
 ## Bundle-gap
 Checked against `bundle/concepts/docs-en-workflows.md`
 (`source_sha 363819ed9ec325275ca22023f6bb6b98fbbf6fcc12db0667478c5207e117751b`,
 timestamp 2026-08-15), plus `docs-en-agents.md` and `docs-en-sub-agents.md`.
 **Confirmed by the bundle:**
 - The **2.1.154 floor** for dynamic workflows, on paid plans.
 - The **main-context relief mechanism**: "A workflow script holds the loop, the branching,
  and the intermediate results itself, so Claude's context holds only the final answer"
  (`:38`); "Intermediate results stay in script variables instead of landing in Claude's
  context" (`:316`). V2's mechanism claim is sound — it is the *magnitude* that is unmeasured.
 - **Per-stage model routing** (`:375`–`:378`).
 - **Cost caps**: agent caps, size guideline, and a `Large workflow` warning above 25 agents
  or 1.5M projected tokens (`:368`).
 **Where V2 overstates, and the bundle wins:**
 - **"tokenbudsjetter" (token budgets).** The page documents *caps and size guidelines*, not
  a token budget — `budget` returns 0 hits on the page. V2's warning to "set an explicit
  token budget" has no documented primitive to point at in the official docs.
 - **"automatisk retry".** 0 hits. Not documented on the page.
 **Limitation V2 omits, material for this repo:** resumability is **session-scoped** —
 "Resume works within the same Claude Code session. If you exit Claude Code while a workflow
 is running, the next session starts the workflow fresh" (`:360`). Further, stopping mid
 fan-out re-runs every agent that started after the stopped one (`:354`). Voyage is
 explicitly a *multi-session* tool, so resumability is a weaker argument here than V2 implies.
 **Gap proper (feedback toward wiki v1.0):**
 1. The mirrored docs describe no programmatic **token-budget** primitive for workflows,
   while the Workflow tool's own runtime surface does expose a budget derived from an
   operator token directive. If that surface is real and stable, it is missing from the
   mirror. Flagged as a discrepancy, not asserted as a doc error.
 2. Nothing in the mirrored pages quantifies what a workflow's **final return** costs the
   orchestrating session. The relief mechanism is described qualitatively only, so a
   consumer cannot size the benefit from the docs alone — which is precisely why the
   measurement below cannot be replaced by reading.
 ---
 ## Recommended order (operator decides; nothing started)
 1. **Fail-closed coordinator (tiltak 2c)** — smallest surface, highest value, no
   dependencies, and it closes a defect rather than adding a feature.
 2. **Measure Δ main-context on the `/trekplan` Phase-5 swarm (tiltak 1)** — Arm A vs Arm B
   against the pre-registered bars in `docs/T1-cc26-delegated-orchestration.md`. The port
   decision follows the number; it does not precede it.
 3. **`gemini-bridge` fix-or-drop (tiltak 5)** — already analysed, currently broken, cheap.
 **Deferred with reasons:** tiltak 3 (already open operator decision 1) · tiltak 4 (requires
 a `brief_version` bump against a public contract) · tiltak 5's opus/synthesis items
 (decided; operator-pinned) · tiltak 6 (harness-level, not repo code).
 ---
 ## Verification log
 | Claim | How verified |
 |---|---|
 | `/trekreview` Workflow port exists and was measured | `commands/trekreview.md:270`; `docs/T2-bakeoff-results.md` §Full run |
 | +4.4% is **subagent** tokens, main context uncounted | `docs/T2-bakeoff-results.md:154` (verbatim) |
 | Δ main-context ≈ 0 applies to Phase 7 only | `docs/T1-synthesis-poc-results.md:1`, `:100` |
 | Measurement design already exists | `docs/T1-cc26-delegated-orchestration.md:118`, `:125` |
 | Reviewers are not forks | `grep subagent_type\|fork commands/ agents/` → only `Explore` at `commands/trekplan.md:501` |
 | Plans/commit messages not fed to reviewers | Phase 5 input list; `commands/trekreview.md:132`, `:181` |
 | Verdict is computed from survivors only | `agents/review-coordinator.md:87`, `:101`, `:132` |
 | Same rule codified deterministically in `lib/` | `lib/review/coordinator-contract.mjs:184`–`:193`, `:206`; existing test `tests/lib/coordinator-contract.test.mjs` |
 | No empty-return handling in Phase 5 | grep → 0, positive control `Phase 5`=7, `reviewer`=24 |
 | Existing fail-closed branch on schema failure | `commands/trekreview.md`, Phase 5 bounded-retry paragraph |
 | `--quick` does not bypass 2.2 | `commands/trekbrief.md:126`, `:472`, `:474` |
 | No DoD fields in trekbrief | grep → 0, positive control `Success Criteri`=5 |
 | 24/24 agents pinned opus, by operator | file count 24/24; `git log 40d8742` |
 | gemini-bridge already flagged for drop | `docs/voyage-vs-cc-balance-analysis.md:81`, `:121`, `:149` |
 | No LSP usage | `grep -rn LSP commands/ agents/ lib/ docs/` → 0 |
 | Workflow floor, relief mechanism, model routing, caps | `bundle/concepts/docs-en-workflows.md:19`, `:38`, `:316`, `:375`, `:368` |
 | No documented token budget or retry in the bundle | `grep -i budget\|retry` on that page → 0 relevant |
 | Resume is session-scoped | `bundle/concepts/docs-en-workflows.md:354`, `:360` |
 | Fork inherits full conversation context | `bundle/concepts/docs-en-agents.md:37` |
 **Not verified / stated as unverified:** V2's underlying video-corpus claims (the StrongDM
 bad-merge case, the 5.5-hour bake-off, the "$200 plan in 30 minutes" workflow) were not
 independently checked — they are cited here as V2's evidence, not as this repo's findings.
--- a/docs/T1-cc26-delegated-orchestration.md
+++ b/docs/T1-cc26-delegated-orchestration.md
@ -135,17 +135,6 @@ reused.
 ## 6. Cheaper PoC (audit-recommended, preferred over the full bake-off)
 > **STATUS: RUN AND DECLINED (S12, 2026-06-18) — do not re-order as new work.**
 > Result: **Δ main-context (faithful flow) = 0.0%**, NEGATIVE against the 15% keep-floor.
 > Full method, fixtures and BASE-sweep: `docs/T1-synthesis-poc-results.md`;
 > reproduce with `node scripts/synthesis-measure.mjs`.
 > The finding is **structural, not stochastic**, so re-running it cannot change the number:
 > Phase 5 spawns the exploration swarm **foreground** (`commands/trekplan.md:158,338-341` —
 > "foreground is the only mode as of v2.4.0"), so the 6-10 outputs are **already resident in
 > main** before Phase 7 begins. Delegating only the Phase-7 digest returns *on top of* them
 > and evicts nothing. A saving is reachable only by also moving Phase-5 delivery off-main —
 > which is §5, not §6. `agents/synthesis-agent.md` therefore ships **dormant**.
 Per `subagent-delegation-audit.md` §Recommendation, the lowest-risk way to test the delegation
 *premise* is **not** a wholesale orchestrator rewrite but **one narrow synthesis-agent**
 (intervention #1/#3): delegate only trekplan Phase 7 synthesis (the heaviest single inline read —
@ -180,17 +169,5 @@ promising orchestration-substrate question and is untouched by this.
   only matters if a future nested pipeline approaches 5.
 2. Classifier behaviour for a *purposeful* swarm under `auto`/`bypassPermissions` is unverified
   (F4) — must be in the §5/§6 test matrix before any delegated spawn ships to a headless path.
-3. ~~The §6 synthesis-agent PoC and the §5 full bake-off are both designed but unbuilt.~~
+3. The §6 synthesis-agent PoC and the §5 full bake-off are both **designed but unbuilt** — ready
-   **CORRECTED 2026-09-03 (S95).** This item was written in S7 and falsified the same afternoon
+   to run if CC-26 is greenlit toward delegation.
   by S12: the **§6 PoC was RUN and DECLINED** (`docs/T1-synthesis-poc-results.md` — Δ_faithful
   = 0.0%, NEGATIVE). Only the **§5 full bake-off** remains unbuilt. The stale wording is what
   caused the §6 PoC to be re-ordered as new work on 2026-09-02; it is struck rather than
   deleted so the re-order has a traceable cause.
 4. **Environment premise updated (measured 2026-09-03, CC 2.1.259).** S12 recorded the live run
   as environment-blocked for want of an `ANTHROPIC_API_KEY`. That half no longer holds:
   `claude -p --output-format stream-json` runs on subscription auth with no API key and emits
   per-turn `usage` plus a `subagent_stats` block (spawned / max_depth / refused) that did not
   exist in June — a ready-made classifier-interference counter for §5's feasibility guard.
   S12's *other* blocker stands unmeasured: the installed plugin is the cache copy, so a
   repo-local agent is invisible to `claude -p`. This changes nothing about the §6 verdict,
   which is structural (see §6 status), and only lowers the cost of §5.
--- a/docs/agent-return-channel-defect.md
+++ b/docs/agent-return-channel-defect.md
@ -167,18 +167,10 @@ arm produces correct final text at 66 lines and at 3730 lines alike, and only
 delivery fails, identically at both. The 38 recovered findings were re-used
 instead of re-run.
-**A second external measurement closes the other half (akashic-intelligence
+What it does **not** close: their four cells were all named, so the *returning*
-S27, commit `f168630`).** The gap left above was that all four S25 cells were
+(unnamed) arm still has no measurement above 66 lines. That a plain subagent
-named, so the *returning* (unnamed) arm had no measurement above 66 lines.
+returns a 21 KB result at that scale is inferred, not observed. Stated as
-S27 supplies one: denominator **2 of 2 unnamed agents**, against a 4370-line
+inferred.
 plan; both returned, 30449 B and 10989 B, both valid JSON. The unnamed arm
 therefore returns at full scale as observed fact, not as inference.
 Two caveats, kept at the strength akashic itself stated them. The measurement
 was taken by the repo that owns the finding, not by an independent third party.
 And byte-identity between the returned string and the file on disk was not
 proven — what is established is that a well-formed result of that size arrived,
 not that it arrived unaltered.
 Their PONG control also carries the same lesson as cell 1 above, in a third
 repo: S25 reported that agent as having "gone idle without sending PONG". It
--- a/docs/architecture.md
+++ b/docs/architecture.md
@ -98,7 +98,7 @@ Which native Claude Code primitive each pipeline step runs on today, and the alt
 | **continue** | Inline reads `.session-state.local.json` → zero-confirm resume | CC `--resume` (transcript replay, not typed work-state → insufficient) |
 | **cross-cutting** | 8 hook scripts: `pre-bash` + `pre-write` guards, `pre-agent-cap` loop-bound enforcement, `post-bash` stats, `session-title`, `pre-`/`post-compact` flush, **`Stop`→OTEL** export | — |
-¹ MCP per research agent: `docs-researcher` → Microsoft Learn + Tavily · `community-`/`security-`/`contrarian-researcher` → Tavily (+ WebSearch/WebFetch). Graceful degradation when an MCP server is absent.
+¹ MCP per research agent: `docs-researcher` → Microsoft Learn + Tavily · `community-`/`security-`/`contrarian-researcher` → Tavily (+ WebSearch/WebFetch) · `gemini-bridge` → Gemini Deep Research MCP. Graceful degradation when an MCP server is absent.
 **Legend:** *production* = wired and active · *dormant* = shipped but wired to nothing (`synthesis-agent`) · *opt-in* = behind a flag (`--workflow`) · *not wired* = considered, deferred (delegated orchestrator, the `AskUserQuestion` brief engine).
--- a/docs/command-modes.md
+++ b/docs/command-modes.md
@ -20,13 +20,13 @@ Always interactive. Phase 3 is a section-driven completeness loop (no hard cap o
 | _(default)_ | Interview + research (local + external) + synthesis + brief (foreground) |
 | `--project <dir>` | Write brief to `{dir}/research/{NN}-{slug}.md` (auto-incremented) |
 | `--quick` | Interview (short) + inline research (no agent swarm) |
-| `--local` | Only codebase analysis agents (skip external research) |
+| `--local` | Only codebase analysis agents (skip external + Gemini) |
 | `--external` | Only external research agents (skip codebase analysis) |
 | `--fg` | No-op alias (foreground is default since v2.4.0) |
 | `--gates {true\|false}` | (v3.4.0) Boolean autonomy-gate flag; present → gating on. Policy (`gates_mode`) detailed under `## Autonomy mode` in `docs/operations.md`. |
 | `--min-brief-version <ver>` | (S18) Warn — never block — if an attached `--project` brief declares a version below `<ver>` (e.g. `2.2`), i.e. sidesteps framing enforcement |
 | `--profile <name>` | (v4.1.0) Model profile for the research phase. |
-| `--engine {swarm\|deep-research}` | (deep-research-engine) Opt-in external-research engine; `deep-research` delegates the external phase to Claude Code's built-in `/deep-research` workflow, which only works on `2.1.154 <= CC < 2.1.218` — from **2.1.218** `/deep-research` is operator-invoked only (the Skill tool refuses with `disable-model-invocation`), so the engine always falls back to `swarm`. Never hard-fails. Default `swarm`. |
+| `--engine {swarm\|deep-research}` | (deep-research-engine) Opt-in external-research engine; `deep-research` delegates the external phase to Claude Code's built-in `/deep-research` workflow (CC 2.1.154+), falls back to `swarm`. Default `swarm`. |
 Flags combine: `--project <dir> --local`, `--external --quick`.
--- a/lib/review/coordinator-contract.mjs
+++ b/lib/review/coordinator-contract.mjs
@ -19,10 +19,8 @@
 // What IS implemented, purely: Pass 1 (triplet dedup → highest-severity-wins
 // survivor + conformance tiebreak + detail concat + raised_by provenance),
 // Pass 2 succinctness + actionability-presence, Pass 3 reasonableness
-// (citation / unknown-rule_key suppression, severity-mismatch correction),
+// (citation / unknown-rule_key drop, severity-mismatch correction), Pass 4
-// Pass 4 verdict thresholds — fail-closed: a suppression that did not REFUTE
+// verdict thresholds. No LLM, no network, no time, no randomness.
 // the finding, and a reviewer that never reported, forbid ALLOW (see
 // classifySuppression). No LLM, no network, no time, no randomness.
 //
 // Reuses: SEVERITY_VALUES / RULE_KEYS / getRule (rule-catalogue.mjs),
 // computeFindingId (finding-id.mjs, triplet), validateFindings
@ -37,71 +35,6 @@ import { validateFindings } from './findings-schema.mjs';
 export const JUDGE_TITLE_MAX = 100;
 export const JUDGE_DETAIL_MAX = 800;
 // ---- Suppression classification (fail-closed) --------------------------------
 //
 // A removal is `dropped` ONLY when the test refuted the finding as a claim
 // about this codebase. Every other removal is `unverified`: the coordinator
 // took the finding out of the count without ever establishing it was unreal,
 // so it may not be spent as evidence of a clean review.
 /**
 * Reasons that REFUTE. `no-citation` is the only one this deterministic subset
 * can emit: a finding whose `file` is empty or whose `line` is negative names
 * no location, so it makes no checkable claim at all
 * (agents/review-coordinator.md Pass 3 — "Speculative 'code might break
 * somewhere' findings have no anchor").
 *
 * `accuracy:refuted` (Pass 2 Accuracy — a citation escaping the repo root) and
 * `file-existence:refuted` (Pass 3 — absent from both working tree and diff)
 * are emitted by the LLM coordinator, whose fs/judgement branches this module
 * excludes. They are declared here anyway: the vocabulary is owned in one
 * place so prose and lib cannot drift.
 */
 export const REFUTING_REASONS = Object.freeze(new Set([
  'no-citation',
  'accuracy:refuted',
  'file-existence:refuted',
 ]));
 /**
 * The reason vocabulary on the unverified side. `file-existence:indeterminate`
 * is emitted by the LLM coordinator's Pass 3 (which runs the fs Glob this
 * module deliberately excludes); the vocabulary is owned here so prose and lib
 * cannot drift.
 */
 export const UNVERIFIED_REASONS = Object.freeze([
  'succinctness:title',
  'succinctness:detail',
  'actionability:empty',
  'unknown-rule_key',
  'file-existence:indeterminate',
 ]);
 /**
 * Classify a suppression reason. Anything not declared refuting is
 * `unverified` — the default is fail-CLOSED, so a reason introduced later
 * without a decision cannot silently move the verdict toward ALLOW.
 * @param {string} reason
 * @returns {'refuted'|'unverified'}
 */
 export function classifySuppression(reason) {
  return REFUTING_REASONS.has(reason) ? 'refuted' : 'unverified';
 }
 /**
 * Tag a finding with its suppression reason and route it to the refuted
 * (`dropped`) or the `unverified` bucket.
 * @param {object} finding
 * @param {string} reason
 * @param {object[]} dropped
 * @param {object[]} unverified
 */
 function suppress(finding, reason, dropped, unverified) {
  const tagged = { ...finding, suppressed_reason: reason };
  if (classifySuppression(reason) === 'refuted') dropped.push(tagged);
  else unverified.push(tagged);
 }
 /**
 * Catalogue-tier rank of a severity: lower number = higher severity.
 * BLOCKER=0 … SUGGESTION=3; an unknown severity ranks last.
@ -189,18 +122,12 @@ export function dedupByTriplet(findings) {
 * (title > 100 or detail > 800 chars) and actionability (recommended_action,
 * when present, must be a non-empty string). The imperative-verb test is
 * excluded (LLM judgement).
 *
 * Both tests read a `.length`; neither examines the claim, so neither can
 * establish the finding is unreal. Both therefore route to `unverified`.
 * `dropped` stays in the signature for the refuting Pass-2 filter this subset
 * excludes (Accuracy: a path-traversal escape IS a refutation).
 * @param {object[]} findings
- * @returns {{ kept: object[], dropped: object[], unverified: object[] }}
+ * @returns {{ kept: object[], dropped: object[] }}
 */
 export function judgeFilter(findings) {
  const kept = [];
  const dropped = [];
  const unverified = [];
  for (const f of findings) {
    const titleLen = (f.title ?? '').length;
    const detailLen = (f.detail ?? '').length;
@ -211,38 +138,31 @@ export function judgeFilter(findings) {
      (typeof f.recommended_action !== 'string' || f.recommended_action.trim().length === 0)) {
      reason = 'actionability:empty';
    }
-    if (reason) suppress(f, reason, dropped, unverified);
+    if (reason) dropped.push({ ...f, suppressed_reason: reason });
    else kept.push(f);
  }
-  return { kept, dropped, unverified };
+  return { kept, dropped };
 }
 /**
 * Pass 3 — Cloudflare reasonableness (deterministic subset): drop findings
 * with no citation (empty file / line < 0) or an unknown rule_key; CORRECT a
 * severity that does not match the catalogue tier (a correction, not a drop).
- * The fs file-existence glob is excluded (I/O) — its indeterminate branch is
+ * The fs file-existence glob is excluded (I/O).
 * prose-side, tokenised as `file-existence:indeterminate`.
 *
 * `no-citation` REFUTES (the finding names no location, so it makes no
 * checkable claim) and is dropped. `unknown-rule_key` does not: an ad-hoc key
 * is a real defect wearing the wrong label — v5.1.1 high-effort mode already
 * KEEPS these, normalised to PLAN_EXECUTE_DRIFT — so it routes to `unverified`.
 * @param {object[]} findings
- * @returns {{ kept: object[], dropped: object[], unverified: object[] }}
+ * @returns {{ kept: object[], dropped: object[] }}
 */
 export function reasonablenessFilter(findings) {
  const kept = [];
  const dropped = [];
  const unverified = [];
  for (const f of findings) {
    if (typeof f.file !== 'string' || f.file.length === 0 ||
      (typeof f.line === 'number' && f.line < 0)) {
-      suppress(f, 'no-citation', dropped, unverified);
+      dropped.push({ ...f, suppressed_reason: 'no-citation' });
      continue;
    }
    if (!RULE_KEYS.has(f.rule_key)) {
-      suppress(f, 'unknown-rule_key', dropped, unverified);
+      dropped.push({ ...f, suppressed_reason: 'unknown-rule_key' });
      continue;
    }
    const rule = getRule(f.rule_key);
@ -252,114 +172,44 @@ export function reasonablenessFilter(findings) {
      kept.push(f);
    }
  }
-  return { kept, dropped, unverified };
+  return { kept, dropped };
 }
 /**
 * Pass 4 — compute the verdict from severity counts (after dedup + filtering).
 * BLOCKER ≥ 1 → BLOCK; else MAJOR ≥ 1 → WARN; else ALLOW.
 *
 * FAIL-CLOSED: ALLOW additionally requires that nothing is `unverified` and
 * that every expected reviewer reported. Neither ever RAISES a verdict — the
 * severity thresholds are untouched — they only forbid the clean one, so the
 * worst case of a false unverified is WARN plus a stated reason, never a
 * silent pass. Unverified findings are NOT counted into a severity tier: their
 * severity is reviewer-asserted and was never substantiated.
 *
 * @param {object[]} findings
- * @param {{ unverified?: object[], missingReviewers?: string[], unattributablePayloads?: number }} [options]
+ * @returns {{ verdict: 'BLOCK'|'WARN'|'ALLOW', counts: Record<string, number> }}
 * @returns {{ verdict: 'BLOCK'|'WARN'|'ALLOW', counts: Record<string, number>, allow_blocked_by: string[] }}
 */
-export function computeVerdict(findings, options = {}) {
+export function computeVerdict(findings) {
  const counts = { BLOCKER: 0, MAJOR: 0, MINOR: 0, SUGGESTION: 0 };
  for (const f of findings) {
    if (counts[f.severity] !== undefined) counts[f.severity] += 1;
  }
  const unverified = options.unverified ?? [];
  const missingReviewers = options.missingReviewers ?? [];
  const unattributablePayloads = options.unattributablePayloads ?? 0;
  const allow_blocked_by = [];
  const byReason = new Map();
  for (const f of unverified) {
    const reason = f?.suppressed_reason ?? 'unspecified';
    byReason.set(reason, (byReason.get(reason) ?? 0) + 1);
  }
  for (const [reason, n] of byReason) allow_blocked_by.push(`unverified:${reason} (${n})`);
  for (const r of missingReviewers) allow_blocked_by.push(`missing-reviewer:${r}`);
  if (unattributablePayloads > 0) allow_blocked_by.push(`unattributable-payload (${unattributablePayloads})`);
  let verdict;
  if (counts.BLOCKER >= 1) verdict = 'BLOCK';
  else if (counts.MAJOR >= 1) verdict = 'WARN';
  else if (allow_blocked_by.length > 0) verdict = 'WARN';
  else verdict = 'ALLOW';
-  return { verdict, counts, allow_blocked_by };
+  return { verdict, counts };
 }
 /**
 * Run the full deterministic contract: ingest → Pass 1 → Pass 2 → Pass 3 → Pass 4.
 *
 * `options.expectedReviewers` names the reviewers this review was supposed to
 * hear from. A reviewer that is absent from the payloads, or whose named
 * payload failed schema validation and was thrown away at ingest, lands in
 * `missing_reviewers`; a payload that failed schema WITHOUT a reviewer name is
 * counted in `unattributable_payloads` instead. Either forbids ALLOW: an
 * unread reviewer is an absent one, and zero findings from a silent reviewer
 * must not read like zero findings from a clean diff.
 *
 * `suppressed` stays the UNION of `dropped` (refuted) and `unverified` so
 * existing consumers keep their meaning; `unverified` is the subset that
 * forbids ALLOW. Do not iterate both and count twice.
 *
 * @param {Array<{reviewer?: string, findings: object[]}>} reviewerPayloads
- * @param {{ expectedReviewers?: string[] }} [options]
+ * @returns {{ verdict: string, counts: Record<string, number>, findings: object[], suppressed: object[], skipped: object[] }}
 * @returns {{ verdict: string, counts: Record<string, number>, findings: object[], suppressed: object[], unverified: object[], skipped: object[], missing_reviewers: string[], unattributable_payloads: number, allow_blocked_by: string[] }}
 */
-export function runContract(reviewerPayloads, options = {}) {
+export function runContract(reviewerPayloads) {
  const { findings: ingested, skipped } = ingest(reviewerPayloads);
  const deduped = dedupByTriplet(ingested);
  const judged = judgeFilter(deduped);
  const reasoned = reasonablenessFilter(judged.kept);
-  const unverified = [...judged.unverified, ...reasoned.unverified];
+  const { verdict, counts } = computeVerdict(reasoned.kept);
  // A reviewer counts as REPORTED only when a payload carrying its name
  // validated. `validateFindings` merely warns on a missing `reviewer`, so a
  // payload can fail schema anonymously: that is an unattributable payload, not
  // a reviewer called "unnamed reviewer". Naming one would invent an agent
  // nobody launched, and would double-count with expectedReviewers when the two
  // are in fact the same failure.
  const skippedNames = new Set();
  let unattributable_payloads = 0;
  for (const s of skipped) {
    if (typeof s.reviewer === 'string' && s.reviewer.length > 0) skippedNames.add(s.reviewer);
    else unattributable_payloads += 1;
  }
  const reported = new Set();
  for (const payload of reviewerPayloads) {
    const name = payload?.reviewer;
    if (typeof name === 'string' && name.length > 0 && !skippedNames.has(name)) reported.add(name);
  }
  const missing_reviewers = [...skippedNames];
  for (const r of options.expectedReviewers ?? []) {
    if (!reported.has(r) && !missing_reviewers.includes(r)) missing_reviewers.push(r);
  }
  const { verdict, counts, allow_blocked_by } = computeVerdict(reasoned.kept, {
    unverified,
    missingReviewers: missing_reviewers,
    unattributablePayloads: unattributable_payloads,
  });
  return {
    verdict,
    counts,
    findings: reasoned.kept,
-    suppressed: [...judged.dropped, ...judged.unverified, ...reasoned.dropped, ...reasoned.unverified],
+    suppressed: [...judged.dropped, ...reasoned.dropped],
    unverified,
    skipped,
    missing_reviewers,
    unattributable_payloads,
    allow_blocked_by,
  };
 }
--- a/package.json
+++ b/package.json
@ -1,7 +1,6 @@
 {
  "name": "voyage",
-  "version": "5.10.1",
+  "version": "5.10.0",
  "private": true,
  "description": "Voyage — brief, research, plan, execute, review, continue. Contract-driven Claude Code pipeline. /trekbrief, /trekplan, and /trekreview each end by building a self-contained operator-annotation HTML (scripts/annotate.mjs, modelled on claude-code-100x): select text or click any heading/paragraph/list-item, pick intent (Fiks/Endre/Spørsmål), write comment, copy structured prompt, paste back, Claude revises the .md.",
  "type": "module",
  "engines": {
--- a/settings.json
+++ b/settings.json
@ -14,6 +14,11 @@
    "trekresearch": {
      "defaultMode": "default",
      "maxDimensions": 8,
      "geminiBridge": {
        "enabled": true,
        "pollIntervalSeconds": 30,
        "timeoutMinutes": 25
      },
      "interview": {
        "maxQuestions": 4,
        "typicalQuestions": 3
--- a/templates/research-brief-template.md
+++ b/templates/research-brief-template.md
@ -74,6 +74,15 @@ findings exist.*
 ### Known Issues
 {Common pitfalls, gotchas, real-world problems from community-researcher}
 ## Gemini Second Opinion
 *Independent research result from Gemini Deep Research. Provides a second
 perspective for triangulation. Omit this section if gemini-bridge was not used
 or was unavailable.*
 {Gemini findings reformatted into key findings, sources cited, and areas of
 agreement/disagreement with other agents.}
 ## Synthesis
 *Cross-cutting insights that emerge from combining local and external knowledge.
@ -105,7 +114,7 @@ this section entirely.*
 | # | Source | Type | Quality | Used in |
 |---|--------|------|---------|---------|
-| 1 | {URL or codebase path} | {official / community / codebase} | {high / medium / low} | {dimension name} |
+| 1 | {URL or codebase path} | {official / community / codebase / gemini} | {high / medium / low} | {dimension name} |
 *Quality assessment:*
 - **high** — official documentation, verified codebase analysis, peer-reviewed
--- a/templates/trekbrief-template.md
+++ b/templates/trekbrief-template.md
@ -129,7 +129,7 @@ the codebase and this brief contain sufficient context for planning."}
 - **Required for plan steps:** {which kinds of steps will consume this — e.g.,
  "migration strategy", "library selection", "threat model"}
 - **Confidence needed:** {high | medium | low}
- **Estimated cost:** {quick — inline research | standard — agent swarm | deep — with contrarian}
+- **Estimated cost:** {quick — inline research | standard — agent swarm | deep — with contrarian + gemini}
 - **Scope hint:** {local | external | both}
 ### Topic 2: {Short title}
--- a/tests/commands/trekexecute-parallel-portability.test.mjs
+++ b/tests/commands/trekexecute-parallel-portability.test.mjs
@ -1,349 +0,0 @@
 // tests/commands/trekexecute-parallel-portability.test.mjs
 //
 // Executable tests for the two shell blocks /trekexecute Phase 2.6 Step 2a' and
 // Phase 2.55 Check 2 tell the agent to run. Both blocks wrecked a real voyage on
 // macOS (order 20260831T214411Z-941965142-from-.claude):
 //
 //   Defect 1: `realpath --relative-to` is GNU coreutils. On BSD realpath the
 //     command substitution fails, PROJECT_REL becomes EMPTY, and `mkdir -p
 //     "$wt/"` + `cp ... "$wt//"` both SUCCEED — brief.md/plan.md land at the
 //     worktree root instead of the project relpath. Exit status stays 0; only
 //     file location tells the truth. Every assertion here checks placement.
 //   Defect 2: Check 2 ran `git add {plan-path}` unconditionally. When the
 //     project directory is gitignored (normal — .claude/projects/ is tool-
 //     managed and local-only) the add fails and the plan never reaches HEAD.
 //
 // The blocks are EXTRACTED from commands/trekexecute.md and executed, so the
 // test binds to what an agent actually copies, not to prose about it. Anchors
 // are grep-able strings, never line numbers.
 import { test } from 'node:test';
 import { strict as assert } from 'node:assert';
 import { readFileSync, writeFileSync, mkdirSync, mkdtempSync, rmSync, chmodSync } from 'node:fs';
 import { execFileSync, spawnSync } from 'node:child_process';
 import { tmpdir } from 'node:os';
 import { realpathSync } from 'node:fs';
 import { dirname, join } from 'node:path';
 import { fileURLToPath } from 'node:url';
 const HERE = dirname(fileURLToPath(import.meta.url));
 const ROOT = join(HERE, '..', '..');
 const COMMAND_FILE = join(ROOT, 'commands', 'trekexecute.md');
 const COPY_BLOCK_ANCHOR = "**2a'. Copy gitignored project artifacts";
 const CHECK2_ANCHOR = '### Check 2 —';
 // The pre-fix form, kept ONLY as the negative control for Defect 1.
 const LEGACY_GNU_LINE = 'PROJECT_REL="$(realpath --relative-to="$REPO_ROOT" "$PROJECT_SOURCE")"';
 // --- helpers -------------------------------------------------------------
 /** Extract the first ```bash fence that follows `anchor` in commands/trekexecute.md. */
 function extractBashBlock(anchor) {
  const text = readFileSync(COMMAND_FILE, 'utf8');
  const at = text.indexOf(anchor);
  assert.ok(at >= 0, `anchor not found in trekexecute.md: ${anchor}`);
  const fenceOpen = text.indexOf('```bash', at);
  assert.ok(fenceOpen >= 0, `no bash fence after anchor: ${anchor}`);
  const bodyStart = text.indexOf('\n', fenceOpen) + 1;
  const fenceClose = text.indexOf('```', bodyStart);
  assert.ok(fenceClose > bodyStart, `unterminated bash fence after anchor: ${anchor}`);
  return text.slice(bodyStart, fenceClose);
 }
 /**
 * A PATH directory whose `realpath` behaves like BSD realpath: it rejects every
 * GNU long option and resolves bare paths correctly. Stubbed, never assumed —
 * the machine running the suite may or may not have GNU coreutils.
 */
 function bsdRealpathStubDir() {
  const dir = mkdtempSync(join(tmpdir(), 'trek-bsdstub-'));
  const stub = join(dir, 'realpath');
  writeFileSync(stub, [
    '#!/bin/sh',
    '# BSD realpath stand-in: no GNU long options.',
    'for a in "$@"; do',
    '  case "$a" in',
    '    --*) echo "realpath: illegal option -- -" >&2; exit 1 ;;',
    '  esac',
    'done',
    "exec python3 -c 'import os,sys",
    'for p in sys.argv[1:]: print(os.path.realpath(p))',
    "' \"$@\"",
    '',
  ].join('\n'));
  chmodSync(stub, 0o755);
  return dir;
 }
 /** A PATH directory whose `git check-ignore` dies fatally (128), everything else real. */
 function fatalCheckIgnoreGitStubDir() {
  const realGit = execFileSync('/usr/bin/env', ['sh', '-c', 'command -v git'], { encoding: 'utf8' }).trim();
  assert.ok(realGit, 'git not on PATH — cannot build the check-ignore stub');
  const dir = mkdtempSync(join(tmpdir(), 'trek-gitstub-'));
  const stub = join(dir, 'git');
  writeFileSync(stub, [
    '#!/bin/sh',
    '# `git check-ignore` fatal (128): NOT an answer about ignore status.',
    'for a in "$@"; do',
    '  if [ "$a" = "check-ignore" ]; then',
    '    echo "fatal: simulated check-ignore failure" >&2',
    '    exit 128',
    '  fi',
    'done',
    `exec ${realGit} "$@"`,
    '',
  ].join('\n'));
  chmodSync(stub, 0o755);
  return dir;
 }
 function git(cwd, ...args) {
  return execFileSync('git', args, { cwd, encoding: 'utf8', stdio: ['ignore', 'pipe', 'pipe'] });
 }
 /**
 * A temp repo with a committed file plus a project directory holding brief/plan
 * (+ research unless opts.research === false). `gitignoreProject: true` adds the
 * .gitignore line that makes the project directory invisible to `git add` —
 * and, note, invisible to `git status --porcelain` too, which is exactly why
 * this topology sails through Check 1 and only trips on Check 2.
 */
 function makeRepo(opts = {}) {
  const root = realpathSync(mkdtempSync(join(tmpdir(), 'trek-exec-')));
  git(root, 'init', '-q', '-b', 'main');
  git(root, 'config', 'user.email', 'test@example.invalid');
  git(root, 'config', 'user.name', 'Test');
  writeFileSync(join(root, 'README.md'), '# fixture\n');
  if (opts.gitignoreProject) writeFileSync(join(root, '.gitignore'), '.claude/projects/\n');
  git(root, 'add', 'README.md', ...(opts.gitignoreProject ? ['.gitignore'] : []));
  git(root, 'commit', '-qm', 'init');
  const projectRel = join('.claude', 'projects', 'demo');
  const projectDir = join(root, projectRel);
  mkdirSync(projectDir, { recursive: true });
  writeFileSync(join(projectDir, 'brief.md'), '# brief\n');
  writeFileSync(join(projectDir, 'plan.md'), '# plan\n');
  if (opts.research !== false) {
    mkdirSync(join(projectDir, 'research'), { recursive: true });
    writeFileSync(join(projectDir, 'research', '01-x.md'), '# r\n');
  }
  const worktreeDir = join(root, '.claude', 'trekplan-sessions', 'demo', 'worktrees');
  mkdirSync(join(worktreeDir, 'session-1'), { recursive: true });
  mkdirSync(join(worktreeDir, 'session-2'), { recursive: true });
  return { root, projectRel, projectDir, worktreeDir };
 }
 function runBlock(script, { cwd, env }) {
  return spawnSync('bash', ['-c', script], { cwd, env: { ...process.env, ...env }, encoding: 'utf8' });
 }
 function copyBlockEnv(repo, pathPrefixDir) {
  return {
    REPO_ROOT: repo.root,
    PROJECT_DIR: repo.projectDir,
    WORKTREE_DIR: repo.worktreeDir,
    PATH: `${pathPrefixDir}:${process.env.PATH}`,
  };
 }
 function exists(p) {
  try { readFileSync(p); return true; } catch { return false; }
 }
 // --- Defect 1: portable relpath derivation -------------------------------
 test("2a' — the BSD realpath stub is known-positive: bare path resolves, --relative-to is rejected", () => {
  const stubDir = bsdRealpathStubDir();
  try {
    const bare = runBlock('realpath "$HOME"', { cwd: ROOT, env: { PATH: `${stubDir}:${process.env.PATH}` } });
    assert.equal(bare.status, 0, 'stub must resolve a bare path (proves it can succeed)');
    assert.equal(bare.stdout.trim(), realpathSync(process.env.HOME));
    const gnu = runBlock('realpath --relative-to=/ "$HOME"', { cwd: ROOT, env: { PATH: `${stubDir}:${process.env.PATH}` } });
    assert.notEqual(gnu.status, 0, 'stub must reject the GNU long option');
    assert.match(gnu.stderr, /illegal option/, 'stub must fail the way BSD realpath fails');
    assert.equal(gnu.stdout.trim(), '', 'no stdout — this is what leaves PROJECT_REL empty');
  } finally {
    rmSync(stubDir, { recursive: true, force: true });
  }
 });
 test("2a' — NEGATIVE CONTROL: the GNU form drops plan.md at the worktree ROOT under BSD realpath", () => {
  const repo = makeRepo();
  const stubDir = bsdRealpathStubDir();
  try {
    const legacy = [
      'PROJECT_SOURCE="$(realpath "${PROJECT_DIR}")"',
      LEGACY_GNU_LINE,
      'for wt in "$WORKTREE_DIR"/session-*; do',
      '  [ -d "$wt" ] || continue',
      '  mkdir -p "$wt/$PROJECT_REL"',
      '  cp "$PROJECT_SOURCE"/brief.md "$wt/$PROJECT_REL/"',
      '  cp "$PROJECT_SOURCE"/plan.md  "$wt/$PROJECT_REL/"',
      'done',
    ].join('\n');
    runBlock(legacy, { cwd: repo.root, env: copyBlockEnv(repo, stubDir) });
    const wt = join(repo.worktreeDir, 'session-1');
    assert.equal(exists(join(wt, repo.projectRel, 'plan.md')), false,
      'the broken form must NOT put plan.md at the project relpath');
    assert.equal(exists(join(wt, 'plan.md')), true,
      'the broken form silently drops plan.md at the worktree root — the measured havari');
  } finally {
    rmSync(repo.root, { recursive: true, force: true });
    rmSync(stubDir, { recursive: true, force: true });
  }
 });
 test("2a' — the shipped block copies brief/plan/research to $wt/$PROJECT_REL without GNU realpath", () => {
  const repo = makeRepo();
  const stubDir = bsdRealpathStubDir();
  try {
    const r = runBlock(extractBashBlock(COPY_BLOCK_ANCHOR), { cwd: repo.root, env: copyBlockEnv(repo, stubDir) });
    assert.equal(r.status, 0, `block must succeed without GNU realpath. stderr: ${r.stderr}`);
    for (const s of ['session-1', 'session-2']) {
      const dest = join(repo.worktreeDir, s, repo.projectRel);
      assert.equal(exists(join(dest, 'plan.md')), true, `${s}: plan.md must reach $wt/$PROJECT_REL`);
      assert.equal(exists(join(dest, 'brief.md')), true, `${s}: brief.md must reach $wt/$PROJECT_REL`);
      assert.equal(exists(join(dest, 'research', '01-x.md')), true, `${s}: research/ must reach $wt/$PROJECT_REL`);
      assert.equal(exists(join(repo.worktreeDir, s, 'plan.md')), false,
        `${s}: nothing may land at the worktree root`);
    }
  } finally {
    rmSync(repo.root, { recursive: true, force: true });
    rmSync(stubDir, { recursive: true, force: true });
  }
 });
 test("2a' — a project without research/ is not a failure (block still exits 0)", () => {
  const repo = makeRepo({ research: false });
  const stubDir = bsdRealpathStubDir();
  try {
    const r = runBlock(extractBashBlock(COPY_BLOCK_ANCHOR), { cwd: repo.root, env: copyBlockEnv(repo, stubDir) });
    assert.equal(r.status, 0, `missing research/ must not fail the wave. stderr: ${r.stderr}`);
    assert.equal(exists(join(repo.worktreeDir, 'session-1', repo.projectRel, 'plan.md')), true);
  } finally {
    rmSync(repo.root, { recursive: true, force: true });
    rmSync(stubDir, { recursive: true, force: true });
  }
 });
 test("2a' — an underivable relpath fails LOUDLY instead of dropping files at the worktree root", () => {
  const repo = makeRepo();
  const stubDir = bsdRealpathStubDir();
  const outside = realpathSync(mkdtempSync(join(tmpdir(), 'trek-outside-')));
  try {
    mkdirSync(join(outside, 'p'), { recursive: true });
    writeFileSync(join(outside, 'p', 'brief.md'), 'b');
    writeFileSync(join(outside, 'p', 'plan.md'), 'p');
    const env = { ...copyBlockEnv(repo, stubDir), PROJECT_DIR: join(outside, 'p') };
    const r = runBlock(extractBashBlock(COPY_BLOCK_ANCHOR), { cwd: repo.root, env });
    assert.notEqual(r.status, 0, 'a project outside REPO_ROOT must abort the wave');
    assert.match(r.stderr, /relpath/i, 'the abort must name the cause');
    assert.equal(exists(join(repo.worktreeDir, 'session-1', 'plan.md')), false,
      'nothing may be dropped at the worktree root');
  } finally {
    rmSync(repo.root, { recursive: true, force: true });
    rmSync(outside, { recursive: true, force: true });
    rmSync(stubDir, { recursive: true, force: true });
  }
 });
 test("2a' — the GNU-only form is gone from the block agents copy", () => {
  const block = extractBashBlock(COPY_BLOCK_ANCHOR);
  assert.equal(block.includes('--relative-to'), false,
    'the copied block must not contain `realpath --relative-to` (GNU-only)');
  assert.ok(block.includes('os.path.relpath'), 'the copied block must derive the relpath portably');
 });
 test("both copied blocks are ASCII-clean (bash 3.2 dies on a multibyte char under set -u)", () => {
  const nonAscii = (s) => s.split('\n')
    .map((line, i) => [i + 1, line])
    .filter(([, line]) => /[^\x00-\x7F]/.test(line));
  // Known-positive: the detector must actually fire on a multibyte char.
  assert.equal(nonAscii('echo "a — b"').length, 1, 'detector must find an em-dash');
  for (const anchor of [COPY_BLOCK_ANCHOR, CHECK2_ANCHOR]) {
    assert.deepEqual(nonAscii(extractBashBlock(anchor)), [],
      `non-ASCII inside the shell block after ${anchor} (prose outside the fence is fine)`);
  }
 });
 // --- Defect 2: gitignored project directory ------------------------------
 // The block carries the `{plan-path}` placeholder the way every other block in
 // trekexecute.md does. Substitute it exactly as an agent would — never inject
 // PLAN_PATH through the environment: that would supply what the doc must supply
 // itself, and a block that never assigns the variable would still pass.
 function check2Script(planPath) {
  const block = extractBashBlock(CHECK2_ANCHOR);
  assert.ok(block.includes('{plan-path}'),
    'Check 2 block must carry the {plan-path} placeholder for the agent to substitute');
  return block.replace('{plan-path}', planPath);
 }
 function check2Env(repo, pathPrefixDir) {
  const env = { REPO_ROOT: repo.root, PLAN_PATH: '' };
  if (pathPrefixDir) env.PATH = `${pathPrefixDir}:${process.env.PATH}`;
  return env;
 }
 test('Check 2 — gitignored project dir: no commit, no failure, and 2a\' still delivers the plan', () => {
  const repo = makeRepo({ gitignoreProject: true });
  const stubDir = bsdRealpathStubDir();
  try {
    // Known-positive on the premise: the plan file really is ignored here.
    const ci = spawnSync('git', ['check-ignore', '-v', join(repo.projectRel, 'plan.md')],
      { cwd: repo.root, encoding: 'utf8' });
    assert.equal(ci.status, 0, 'fixture premise: the plan file must actually be gitignored');
    const head = git(repo.root, 'rev-parse', 'HEAD').trim();
    const r = runBlock(check2Script(join(repo.projectRel, 'plan.md')),
      { cwd: repo.root, env: check2Env(repo) });
    assert.equal(r.status, 0, `Check 2 must tolerate a gitignored plan file. stderr: ${r.stderr}`);
    assert.equal(git(repo.root, 'rev-parse', 'HEAD').trim(), head,
      'an ignored plan file must NOT be forced into history (origin is a public mirror)');
    const copy = runBlock(extractBashBlock(COPY_BLOCK_ANCHOR), { cwd: repo.root, env: copyBlockEnv(repo, stubDir) });
    assert.equal(copy.status, 0, `copy step must succeed. stderr: ${copy.stderr}`);
    assert.equal(exists(join(repo.worktreeDir, 'session-1', repo.projectRel, 'plan.md')), true,
      'the plan must reach the worktree even though git never tracked it');
  } finally {
    rmSync(repo.root, { recursive: true, force: true });
    rmSync(stubDir, { recursive: true, force: true });
  }
 });
 test('Check 2 — KNOWN-POSITIVE: an untracked, NOT-ignored plan file is still added and committed', () => {
  const repo = makeRepo();
  try {
    const planPath = join(repo.projectRel, 'plan.md');
    const head = git(repo.root, 'rev-parse', 'HEAD').trim();
    const r = runBlock(check2Script(planPath), { cwd: repo.root, env: check2Env(repo) });
    assert.equal(r.status, 0, `Check 2 must succeed on a normal untracked plan. stderr: ${r.stderr}`);
    assert.notEqual(git(repo.root, 'rev-parse', 'HEAD').trim(), head,
      'a trackable plan file must still be committed for worktree visibility');
    const ls = spawnSync('git', ['ls-files', '--error-unmatch', planPath], { cwd: repo.root, encoding: 'utf8' });
    assert.equal(ls.status, 0, 'the plan file must now be tracked');
  } finally {
    rmSync(repo.root, { recursive: true, force: true });
  }
 });
 test('Check 2 — a FATAL git check-ignore (128) is not read as "not ignored"', () => {
  const repo = makeRepo({ gitignoreProject: true });
  const stubDir = fatalCheckIgnoreGitStubDir();
  try {
    const head = git(repo.root, 'rev-parse', 'HEAD').trim();
    const r = runBlock(check2Script(join(repo.projectRel, 'plan.md')),
      { cwd: repo.root, env: check2Env(repo, stubDir) });
    assert.notEqual(r.status, 0, 'a fatal check-ignore must stop, not fall through to git add');
    assert.match(r.stderr, /check-ignore/, 'the stop must name the failing probe');
    assert.equal(git(repo.root, 'rev-parse', 'HEAD').trim(), head, 'no commit may be made on a fatal probe');
  } finally {
    rmSync(repo.root, { recursive: true, force: true });
    rmSync(stubDir, { recursive: true, force: true });
  }
 });
--- a/tests/lib/agent-frontmatter.test.mjs
+++ b/tests/lib/agent-frontmatter.test.mjs
@ -152,18 +152,14 @@ test('no agents/*.md frontmatter contains an <example> block (M4: examples live
  }
 });
-// Floor lowered 34 -> 32 in v5.10.1: `gemini-bridge` (2 examples) was DELETED as an
+test('agent bodies retain at least 34 <example> blocks (M4: relocation moves, never deletes)', () => {
 // agent by operator directive, not relocated. The pin guards M4's "relocation moves,
 // never deletes" invariant, so it must move only when an agent legitimately leaves the
 // inventory — never to absorb examples that went missing from a surviving agent.
 test('agent bodies retain at least 32 <example> blocks (M4: relocation moves, never deletes)', () => {
  let total = 0;
  for (const f of agentFiles) {
    total += (bodyOf(read(`agents/${f}`)).match(/<example>/g) || []).length;
  }
  assert.ok(
-    total >= 32,
+    total >= 34,
-    `expected >= 32 <example> blocks across agent bodies (16 agents x 2), got ${total} ` +
+    `expected >= 34 <example> blocks across agent bodies (17 agents x 2), got ${total} ` +
    `— examples may have been deleted instead of relocated (M4)`,
  );
 });
--- a/tests/lib/coordinator-contract.test.mjs
+++ b/tests/lib/coordinator-contract.test.mjs
@ -6,9 +6,6 @@
 import { test } from 'node:test';
 import { strict as assert } from 'node:assert';
 import { readFileSync } from 'node:fs';
 import { join, dirname } from 'node:path';
 import { fileURLToPath } from 'node:url';
 import {
  severityRank,
  ingest,
@ -16,14 +13,9 @@ import {
  judgeFilter,
  reasonablenessFilter,
  computeVerdict,
  classifySuppression,
  REFUTING_REASONS,
  UNVERIFIED_REASONS,
  runContract,
 } from '../../lib/review/coordinator-contract.mjs';
 const ROOT = join(dirname(fileURLToPath(import.meta.url)), '..', '..');
 // ---- Pass 1 — dedup --------------------------------------------------------
 test('dedupByTriplet — genuine cross-reviewer collapse (identical triplet) → 1, raised_by both', () => {
@ -106,41 +98,29 @@ test('computeVerdict — counts each severity tier', () => {
 // ---- Pass 3 — reasonableness -----------------------------------------------
-test('reasonablenessFilter — citation-less is REFUTED, unknown rule_key is UNVERIFIED, severity mismatch corrected', () => {
+test('reasonablenessFilter — drops unknown rule_key + citation-less, corrects severity mismatch', () => {
  // Contract change (fail-closed): only `no-citation` refutes. An ad-hoc
  // rule_key is a real defect wearing the wrong label — v5.1.1 high-effort mode
  // already keeps those, normalised to PLAN_EXECUTE_DRIFT.
  const r = reasonablenessFilter([
-    { file: 'x.mjs', line: 1, rule_key: 'NOPE_KEY', severity: 'BLOCKER' },        // unknown → unverified
+    { file: 'x.mjs', line: 1, rule_key: 'NOPE_KEY', severity: 'BLOCKER' },        // unknown → drop
    { file: '', line: 1, rule_key: 'MISSING_TEST', severity: 'MAJOR' },           // no file → drop
    { file: 'x.mjs', line: -1, rule_key: 'MISSING_TEST', severity: 'MAJOR' },     // line < 0 → drop
    { file: 'x.mjs', line: 1, rule_key: 'MISSING_TEST', severity: 'MINOR' },      // catalogue is MAJOR → correct, keep
  ]);
  assert.equal(r.kept.length, 1);
-  assert.equal(r.dropped.length, 2);
+  assert.equal(r.dropped.length, 3);
  assert.deepEqual(r.dropped.map((f) => f.suppressed_reason), ['no-citation', 'no-citation']);
  assert.equal(r.unverified.length, 1);
  assert.equal(r.unverified[0].suppressed_reason, 'unknown-rule_key');
  assert.equal(r.kept[0].severity, 'MAJOR');
  assert.equal(r.kept[0].original_severity, 'MINOR');
 });
 // ---- Pass 2 — judge --------------------------------------------------------
-test('judgeFilter — over-long title and empty recommended_action are UNVERIFIED, not dropped', () => {
+test('judgeFilter — drops over-long title and empty recommended_action', () => {
  // Contract change (fail-closed): both implemented Pass 2 tests read a
  // `.length` and never examine the claim, so neither refutes the finding.
  // `dropped` is empty here on purpose — the refuting Pass 2 filter (Accuracy)
  // is the one this deterministic subset excludes.
  const j = judgeFilter([
-    { file: 'x.mjs', line: 1, rule_key: 'MISSING_TEST', severity: 'MAJOR', title: 'x'.repeat(101) },          // too long → unverified
+    { file: 'x.mjs', line: 1, rule_key: 'MISSING_TEST', severity: 'MAJOR', title: 'x'.repeat(101) },          // too long → drop
-    { file: 'x.mjs', line: 2, rule_key: 'MISSING_TEST', severity: 'MAJOR', title: 'ok', recommended_action: '   ' }, // empty action → unverified
+    { file: 'x.mjs', line: 2, rule_key: 'MISSING_TEST', severity: 'MAJOR', title: 'ok', recommended_action: '   ' }, // empty action → drop
    { file: 'x.mjs', line: 3, rule_key: 'MISSING_TEST', severity: 'MAJOR', title: 'ok' },                     // keep (no action field is fine)
  ]);
  assert.equal(j.kept.length, 1);
-  assert.equal(j.dropped.length, 0);
+  assert.equal(j.dropped.length, 2);
  assert.equal(j.unverified.length, 2);
  assert.deepEqual(j.unverified.map((f) => f.suppressed_reason), ['succinctness:title', 'actionability:empty']);
 });
 // ---- ingest ----------------------------------------------------------------
@ -181,155 +161,3 @@ test('runContract — deterministic: identical input yields identical output', (
  ];
  assert.deepEqual(runContract(input), runContract(input));
 });
 // ---- Fail-closed: the `unverified` bucket (ORDRE 834432937) -----------------
 //
 // The defect: a finding REMOVED by Pass 2/Pass 3, and a reviewer whose payload
 // was thrown away or never arrived, are all arithmetically identical to a
 // finding that never existed -- they push the verdict toward ALLOW. Measured
 // before the fix (probe, 2026-09-01): an over-long-title BLOCKER -> ALLOW; a
 // payload with one ad-hoc rule_key -> the whole payload skipped, its valid
 // BLOCKER sibling gone -> ALLOW.
 //
 // The rule under test: a removal is `dropped` ONLY when the test refutes the
 // finding as a claim about this codebase. Every other removal is `unverified`,
 // and a non-empty `unverified` -- or a reviewer that did not report -- forbids
 // ALLOW.
 test('classifySuppression — only no-citation refutes; form and taxonomy failures are unverified', () => {
  assert.equal(classifySuppression('no-citation'), 'refuted',
    'a finding that names no location makes no checkable claim');
  assert.equal(classifySuppression('succinctness:title'), 'unverified');
  assert.equal(classifySuppression('succinctness:detail'), 'unverified');
  assert.equal(classifySuppression('actionability:empty'), 'unverified');
  assert.equal(classifySuppression('unknown-rule_key'), 'unverified');
  assert.equal(classifySuppression('file-existence:indeterminate'), 'unverified');
  assert.equal(classifySuppression('something-nobody-declared'), 'unverified',
    'an unclassified reason must fail CLOSED, not open');
 });
 test('computeVerdict — non-empty unverified forbids ALLOW but never downgrades BLOCK or WARN', () => {
  const u = [{ file: 'x.mjs', line: 1, rule_key: 'MISSING_TEST', severity: 'BLOCKER' }];
  const withUnverified = computeVerdict([], { unverified: u });
  assert.equal(withUnverified.verdict, 'WARN', 'ALLOW is forbidden while anything is unverified');
  assert.deepEqual(withUnverified.counts, { BLOCKER: 0, MAJOR: 0, MINOR: 0, SUGGESTION: 0 },
    'the unverified finding is NOT counted into a severity tier');
  assert.ok(withUnverified.allow_blocked_by.length > 0);
  assert.equal(computeVerdict([{ severity: 'BLOCKER' }], { unverified: u }).verdict, 'BLOCK',
    'BLOCK stands regardless of the unverified bucket');
  assert.equal(computeVerdict([{ severity: 'MAJOR' }], { unverified: u }).verdict, 'WARN');
  assert.equal(computeVerdict([], { unverified: [] }).verdict, 'ALLOW',
    'known-positive control: an empty unverified bucket still allows ALLOW');
 });
 test('computeVerdict — a reviewer that did not report forbids ALLOW', () => {
  const r = computeVerdict([], { missingReviewers: ['brief-conformance-reviewer'] });
  assert.equal(r.verdict, 'WARN');
  assert.ok(r.allow_blocked_by.some((x) => x.includes('brief-conformance-reviewer')));
 });
 test('runContract — a BLOCKER dropped for an over-long title cannot yield ALLOW', () => {
  // Pass 2 succinctness reads `.length`. It never examines the claim, so it
  // cannot establish the finding is unreal -- it is unverified, not refuted.
  const result = runContract([
    { reviewer: 'code-correctness-reviewer', findings: [
      { file: 'lib/auth/jwt.mjs', line: 19, rule_key: 'SECURITY_INJECTION', severity: 'BLOCKER', title: 'x'.repeat(101), detail: 'algo taken from the JWT header' },
    ] },
  ]);
  assert.notEqual(result.verdict, 'ALLOW', 'an unsubstantiated BLOCKER must never clear the review');
  assert.equal(result.findings.length, 0, 'it is still not a kept finding');
  assert.equal(result.unverified.length, 1);
  assert.equal(result.unverified[0].suppressed_reason, 'succinctness:title');
  assert.equal(result.suppressed.length, 1, 'suppressed stays the union of dropped + unverified');
 });
 test('runContract — a schema-invalid payload cannot yield ALLOW (an unread reviewer is an absent one)', () => {
  // Measured: one ad-hoc rule_key invalidates the WHOLE payload at ingest, so a
  // valid BLOCKER sibling disappears with it. That must not read as "clean".
  const result = runContract([
    { reviewer: 'code-correctness-reviewer', findings: [
      { file: 'lib/auth/jwt.mjs', line: 19, rule_key: 'SECURITY_INJECTION', severity: 'BLOCKER', title: 'real' },
      { file: 'x.mjs', line: 1, rule_key: 'NOPE_KEY', severity: 'MINOR', title: 'ad-hoc key' },
    ] },
  ]);
  assert.equal(result.skipped.length, 1);
  assert.notEqual(result.verdict, 'ALLOW');
  assert.ok(result.allow_blocked_by.some((x) => x.includes('code-correctness-reviewer')));
 });
 test('runContract — a reviewer named in expectedReviewers that never reported cannot yield ALLOW', () => {
  const result = runContract(
    [{ reviewer: 'code-correctness-reviewer', findings: [] }],
    { expectedReviewers: ['code-correctness-reviewer', 'brief-conformance-reviewer'] },
  );
  assert.deepEqual(result.missing_reviewers, ['brief-conformance-reviewer']);
  assert.notEqual(result.verdict, 'ALLOW');
 });
 test('runContract — known-positive control: every reviewer reported, nothing suppressed → ALLOW', () => {
  // Proves ALLOW is still REACHABLE. Without this, "no ALLOW" is not a
  // fail-closed contract, only a broken one.
  const result = runContract(
    [
      { reviewer: 'code-correctness-reviewer', findings: [
        { file: 'a.mjs', line: 1, rule_key: 'MISSING_ERROR_HANDLING', severity: 'MINOR', title: 'unguarded await', recommended_action: 'Wrap the await in a try/catch.' },
      ] },
      { reviewer: 'brief-conformance-reviewer', findings: [] },
    ],
    { expectedReviewers: ['code-correctness-reviewer', 'brief-conformance-reviewer'] },
  );
  assert.equal(result.verdict, 'ALLOW');
  assert.equal(result.unverified.length, 0);
  assert.deepEqual(result.missing_reviewers, []);
  assert.deepEqual(result.allow_blocked_by, []);
 });
 test('classifySuppression — the refuting reasons the LLM coordinator emits are declared here too', () => {
  // agents/review-coordinator.md Pass 2 "Accuracy" and Pass 3 "Non-existent
  // file" DO refute (a citation outside the repo root, a file absent from both
  // tree and diff). Both are fs/judgement branches this deterministic subset
  // excludes, but the vocabulary is owned here so prose and lib cannot drift.
  assert.equal(classifySuppression('accuracy:refuted'), 'refuted');
  assert.equal(classifySuppression('file-existence:refuted'), 'refuted');
  assert.equal(classifySuppression('file-existence:indeterminate'), 'unverified',
    'unresolvable must never collapse into refuted');
 });
 test('suppression vocabulary — the two sets are disjoint and every reason is documented in the prose', () => {
  const refuting = [...REFUTING_REASONS];
  const overlap = refuting.filter((r) => UNVERIFIED_REASONS.includes(r));
  assert.deepEqual(overlap, [], 'a reason cannot be both refuting and unverified');
  const prose = readFileSync(join(ROOT, 'agents/review-coordinator.md'), 'utf-8');
  assert.ok(prose.includes('review-coordinator'), 'known-positive control: the prose file loaded');
  for (const reason of [...refuting, ...UNVERIFIED_REASONS]) {
    assert.ok(prose.includes(reason),
      `reason "${reason}" is declared in the lib but never documented in agents/review-coordinator.md`);
  }
 });
 test('runContract — an anonymous invalid payload is unattributable, not a reviewer named "unnamed reviewer"', () => {
  // `validateFindings` only WARNS on a missing `reviewer`, so a payload can
  // fail schema while carrying no name. Reporting it as a reviewer name
  // invents an agent nobody launched, and double-counts with expectedReviewers
  // when they are in fact the same failure.
  const result = runContract(
    [{ findings: [{ file: 'x.mjs', line: 1, rule_key: 'NOPE', severity: 'MAJOR' }] }],
    { expectedReviewers: ['code-correctness-reviewer'] },
  );
  assert.deepEqual(result.missing_reviewers, ['code-correctness-reviewer'],
    'missing_reviewers carries real names only');
  assert.equal(result.unattributable_payloads, 1);
  assert.ok(result.allow_blocked_by.some((x) => x.startsWith('unattributable-payload')));
  assert.ok(!result.allow_blocked_by.some((x) => x.includes('unnamed reviewer')));
 });
 test('runContract — an anonymous invalid payload forbids ALLOW on its own, with no expectedReviewers', () => {
  // The fail-closed floor must not depend on the caller passing an expected
  // set: without this, dropping the name from a payload would restore ALLOW.
  const result = runContract([{ findings: [{ file: 'x.mjs', line: 1, rule_key: 'NOPE', severity: 'MAJOR' }] }]);
  assert.deepEqual(result.missing_reviewers, []);
  assert.equal(result.unattributable_payloads, 1);
  assert.notEqual(result.verdict, 'ALLOW');
 });
--- a/tests/lib/doc-consistency.test.mjs
+++ b/tests/lib/doc-consistency.test.mjs
@ -701,23 +701,6 @@ test('producing commands emit file:// link in final report (operator-UX contract
  }
 });
 test('package.json is marked private — the npm name `voyage` belongs to a third party', () => {
  // D-census 2026-08-20/21: registry.npmjs.org/voyage answers 200, but that is
  // NOT this package — it is "Advanced HTTP Routing System for Node.js"
  // (NEURS/voyage, v0.0.1, 2022, maintainer kevin.martin). This manifest is test
  // tooling for a Claude Code plugin distributed through the marketplace catalogue
  // (`ref: vX.Y.Z`), never through npm. Without `private`, nothing in the manifest
  // stops an accidental `npm publish` against a name someone else owns — a one-way
  // action. Sibling non-package manifests (okr, repo-mailbox, repo-standard) all
  // carry the flag; voyage was the exception. Operator decision (S93): option A.
  const pkg = JSON.parse(read('package.json'));
  assert.equal(
    pkg.private,
    true,
    'package.json must declare "private": true — the npm name `voyage` is owned by a third party',
  );
 });
 test('package.json still has no "npm run render" script (removed in v5.0.1)', () => {
  const pkg = JSON.parse(read('package.json'));
  assert.equal(
@ -1134,57 +1117,6 @@ test('deep-research-engine: --engine is documented + consistent across surfaces'
  );
 });
 test('deep-research-engine: the CC 2.1.218 operator-only ceiling is documented on every surface', () => {
  // CC 2.1.218 changed /deep-research to "start only when invoked manually"; from
  // there the Skill tool refuses a model invocation outright with
  // `disable-model-invocation`. The flag stays (additive opt-in, always degrades to
  // swarm), but no surface may keep promising a path Claude Code has removed, and
  // none may leave `2.1.154+` standing alone as the engine's requirement.
  const SURFACES = ['commands/trekresearch.md', 'docs/command-modes.md', 'README.md'];
  for (const f of SURFACES) {
    const body = read(f);
    assert.ok(
      body.includes('2.1.218'),
      `${f} must document the CC 2.1.218 ceiling for the deep-research engine`,
    );
    assert.ok(
      body.includes('disable-model-invocation'),
      `${f} must name disable-model-invocation as the fallback reason from CC 2.1.218`,
    );
  }
  // The one-line reference rows carry the whole truth on their own line — an
  // operator reading the flag table never scrolls to the command prose.
  const ROWS = [
    ['docs/command-modes.md', /^.*`--engine \{swarm\\\|deep-research\}`.*$/m],
    ['README.md', /^\|\s*\*\*Engine\*\*.*$/m],
  ];
  for (const [f, re] of ROWS) {
    const m = read(f).match(re);
    assert.ok(m, `${f} must still carry the --engine reference row`);
    assert.ok(
      m[0].includes('2.1.218') && m[0].includes('disable-model-invocation'),
      `${f} --engine row must state the real window (2.1.218 + disable-model-invocation)`,
    );
  }
  // Both load-bearing regions of the command itself: the flag bullet and the
  // pre-gate that decides whether the delegation is attempted at all.
  const research = read('commands/trekresearch.md');
  const bullet = research.match(/8\. `--engine <name>`[\s\S]*?Flags can be combined/);
  assert.ok(bullet, 'trekresearch.md must still carry the --engine flag bullet');
  assert.ok(
    bullet[0].includes('2.1.218') && bullet[0].includes('disable-model-invocation'),
    'trekresearch.md --engine bullet must state the real window, not just the 2.1.154 floor',
  );
  const pregate = research.match(/\*\*Coarse pre-gate[\s\S]*?authoritative guard\./);
  assert.ok(pregate, 'trekresearch.md must still carry the deep-research pre-gate');
  assert.ok(
    pregate[0].includes('2.1.218'),
    'the pre-gate must carry an upper ceiling at 2.1.218, not only the 2.1.154 floor',
  );
 });
 // ── STORM bounded loop — env-vars documented across the four surfaces ──────
 // Same cross-doc shape as the --engine pin above. An operator-facing switch
 // documented on one surface is a switch most operators never find; and the
@ -1702,94 +1634,3 @@ test('v5.9: no commands/*.md frontmatter carries a model: key (session inheritan
  assert.deepEqual(offenders, [],
    `command frontmatter must omit model: (orchestrator follows the session model); offenders: ${offenders.join(', ')}`);
 });
 // --- S95 (③): gemini-bridge is DROPPED from the pipeline (operator directive) ---
 //
 // Operator directive, three repetitions: gemini-mcp and voyage:gemini-bridge are
 // never to be used. The drop is permanent, not a wait-for-upgrade — so the pin is
 // on the LIVE pipeline surfaces only. Historical records (CHANGELOG entries,
 // measurement/decision docs) must KEEP their references: they say what a past
 // version did or what was measured then, and rewriting them is the same defect
 // class as bumping a version string inside a measurement doc.
 //
 // `gemini_used` is deliberately NOT banned. It is a vestigial trekresearch stats
 // field, pinned to false by the drop and kept so the observability export schema
 // does not break for existing consumers (see TREKRESEARCH_ALLOWED).
 // Live = anything that can still steer a run. Explicit, so a new file forces a decision.
 const GEMINI_LIVE_SURFACES = [
  'README.md',
  'CLAUDE.md',
  'settings.json',
  'docs/architecture.md',
  'docs/command-modes.md',
  'templates/research-brief-template.md',
  'templates/trekbrief-template.md',
  'tests/fixtures/jsonl-schemas.md',
  ...listMd('commands').map((f) => `commands/${f}`),
  ...listMd('agents').map((f) => `agents/${f}`),
 ];
 // Historical/measurement records that MUST keep their gemini references.
 const GEMINI_HISTORICAL = [
  'CHANGELOG.md',
  'docs/cc-upgrade-2.1.181-decision-matrix.md',
  'docs/voyage-vs-cc-balance-analysis.md',
  'docs/BRIEF-vurdering-v2.md',
  'docs/balance-backlog-plan.md',
  'docs/agent-description-token-trim-brief.md',
  'docs/S22-happy-path-dogfood.md',
  'docs/deep-research-engine-brief.md',
  'examples/02-real-cli/REGENERATED.md',
 ];
 test('S95: agents/gemini-bridge.md is removed from the agent inventory', () => {
  assert.ok(
    !listMd('agents').includes('gemini-bridge.md'),
    'gemini-bridge is dropped by operator directive — agents/gemini-bridge.md must not exist',
  );
 });
 test('S95: no live pipeline surface still names the gemini-bridge agent', () => {
  const offenders = [];
  for (const rel of GEMINI_LIVE_SURFACES) {
    // Everything gemini EXCEPT the vestigial gemini_used stats field.
    const hits = (read(rel).match(/gemini[\w-]*/gi) || []).filter(
      (m) => m.toLowerCase() !== 'gemini_used',
    );
    if (hits.length) offenders.push(`${rel} (${[...new Set(hits)].join(', ')})`);
  }
  assert.deepEqual(
    offenders, [],
    `gemini-bridge is dropped — these live surfaces still reference it: ${offenders.join(' · ')}`,
  );
 });
 test('S95: known-positive control — the historical records still DO carry gemini references', () => {
  // Verifiseringsloven ansikt 4: a scanner that finds nothing anywhere proves
  // nothing. This pins that the same query CAN find, so the empty result above
  // is a measurement and not a broken query.
  const silent = GEMINI_HISTORICAL.filter((rel) => !/gemini/i.test(read(rel)));
  assert.deepEqual(
    silent, [],
    `these historical records lost their gemini references — history must not be rewritten: ${silent.join(', ')}`,
  );
 });
 test('S95: trekplan high-effort Pass 2 degrades explicitly instead of failing', () => {
  const t = read('commands/trekplan.md');
  const section = t.split('\n### High-effort behavior (v5.1.1)\n')[1];
  assert.ok(section, 'trekplan.md must still carry the High-effort behavior sub-section');
  const head = section.split('\n## ')[0];
  assert.match(
    head, /unavailable, skipped/,
    'the dropped Pass 2 must read "unavailable, skipped" explicitly — a removed pass that is silent is indistinguishable from one that failed',
  );
 });
 test('S95: gemini_used survives as a vestigial stats field (export schema unbroken)', () => {
  assert.ok(
    TREKRESEARCH_ALLOWED.has('gemini_used'),
    'gemini_used stays in the trekresearch export allowlist — the agent is dropped, the field is pinned false, and removing it would break existing consumers',
  );
 });