M1: outside-voices advisory panel — primitive + safety rails + shared plan-review resolver

ARCHITECTURE.md

@@ -285,7 +285,7 @@ Templates contain the workflows, tips, and examples that require human judgment.
 | `{{DESIGN_METHODOLOGY}}` | `gen-skill-docs.ts` | Shared design audit methodology for /plan-design-review and /design-review |
 | `{{REVIEW_DASHBOARD}}` | `gen-skill-docs.ts` | Review Readiness Dashboard for /ship pre-flight |
 | `{{TEST_BOOTSTRAP}}` | `gen-skill-docs.ts` | Test framework detection, bootstrap, CI/CD setup for /qa, /ship, /design-review |
-| `{{CODEX_PLAN_REVIEW}}` | `gen-skill-docs.ts` | Optional cross-model plan review (Codex or Claude subagent fallback) for /plan-ceo-review and /plan-eng-review |
+| `{{OUTSIDE_VOICES}}` | `resolvers/outside-voices.ts` | Cross-vendor advisory panel for /plan-ceo-review, /plan-eng-review, /plan-devex-review. External CLI voices run through `bin/gstack-panel` (codex on by default; grok/gemini ship off pending live write-denial canaries); the Fable subagent and a native Claude pass are dispatched by the orchestrator; `bin/gstack-vote` tallies every `<voice>.result.json` into one recommendation. Advisory only: it routes into the existing human gate and never sets the verdict. `{{CODEX_PLAN_REVIEW}}` still resolves as a thin alias for out-of-tree templates. |
 | `{{DESIGN_SETUP}}` | `resolvers/design.ts` | Discovery pattern for `$D` design binary, mirrors `{{BROWSE_SETUP}}` |
 | `{{DESIGN_SHOTGUN_LOOP}}` | `resolvers/design.ts` | Shared comparison board feedback loop for /design-shotgun, /plan-design-review, /design-consultation |
 | `{{UX_PRINCIPLES}}` | `resolvers/design.ts` | User behavioral foundations (scanning, satisficing, goodwill reservoir, trunk test) for /design-html, /design-shotgun, /design-review, /plan-design-review |

CHANGELOG.md

@@ -1,5 +1,50 @@
 # Changelog
 
+## [Unreleased — fork: outside-voices panel M1] - 2026-08-30
+
+**Every plan review now gets a cross-vendor advisory panel: independent second
+opinions from other AI systems, tabulated into one recommendation that can
+never block — the user still decides.**
+
+Fork-local milestone (feat/outside-voices-panel, M1 of 3) on top of the ported
+upstream v1.68.2.0 — VERSION deliberately unchanged: this fork tracks upstream
+release numbers, so fork work ships as CHANGELOG entries, not version bumps.
+
+### Added
+- `bin/gstack-panel` — the invocation primitive for external CLI voices
+  (codex/grok/gemini): sequential-foreground, per-voice 540s timeout ladder
+  bounded by the remaining wall-clock, one `<voice>.result.json` per voice on
+  every path, `--collect` salvage, `--redact-only`, and the full per-voice
+  security pipeline (kill-switches, auth preflight, under-codex guard,
+  first-use per-vendor egress consent, redaction gate, fail-closed egress
+  receipts, read-only sandboxes, `panel_budget_usd` projection).
+- `bin/gstack-vote` — the ONLY tabulation path: filename-bound attribution
+  (only canonical `<voice>.result.json` basenames, each voice at most once),
+  mandatory per-run nonce authentication (an un-nonced ready verdict is
+  demoted, fail-closed), repo-aware finding-location resolution, and the
+  per-voice consensus table (vendor-collapsed median, dissents, appendix).
+- `lib/outside-voices/{registry,vote}.ts` — voice registry with mandatory
+  security fields per adapter, the strict `parseVoiceResult` (schema/enum/
+  size/nonce/spoof validation), and `tallyVoices()` per the design's §5
+  truth table.
+- `{{OUTSIDE_VOICES}}` resolver wired into the ceo/eng/devex plan reviews
+  (replaces the codex-only `{{CODEX_PLAN_REVIEW}}` block) — panel runs
+  default-on as a standard, advisory, non-blocking step.
+- Config keys: `gemini_reviews`, `fable_reviews` (free), `panel_budget_usd`
+  (fail-closed on present-but-empty), `grok_reviews_consent`,
+  `gemini_reviews_consent`; `gstack-config has` now prints the raw stored
+  value for provenance-aware callers.
+- Security/invariant test suites (+~1,000 assertions): nonce, redaction,
+  egress-receipt, sandbox-canary (periodic tier), budget, consent, vote
+  truth-table, and ship-review regression pins.
+
+### Changed
+- Roster safe-default: `grok_reviews`/`gemini_reviews` ship **disabled**
+  (design said all-four-on) until each CLI passes a live write-denial canary
+  — grok's sandbox did not deny writes live; gemini is auth-blocked.
+- gemini sandbox pinned to `--approval-mode plan` (`-s read-only` is invalid
+  against the real CLI).
+
 ## [1.68.2.0] - 2026-08-20
 
 **Revoking a paired agent now revokes everything it holds, and the**

CLAUDE.md

@@ -158,7 +158,7 @@ gstack/
 ├── investigate/     # /investigate skill (systematic root-cause debugging)
 ├── spec/            # /spec skill (five-phase spec → GitHub issue, optional agent spawn, /ship auto-closes)
 ├── retro/           # Retrospective skill (includes /retro global cross-project mode)
-├── bin/             # CLI utilities (gstack-repo-mode, gstack-slug, gstack-config, gstack-wtree, gstack-evidence, gstack-issue-guard, etc.)
+├── bin/             # CLI utilities (gstack-repo-mode, gstack-slug, gstack-config, gstack-wtree, gstack-evidence, gstack-issue-guard, gstack-panel, gstack-vote, etc.)
 ├── document-release/ # /document-release skill (post-ship doc updates + Diataxis coverage map)
 ├── document-generate/ # /document-generate skill (Diataxis doc generator: tutorial/how-to/reference/explanation)
 ├── cso/             # /cso skill (OWASP Top 10 + STRIDE security audit)
@@ -171,7 +171,7 @@ gstack/
 │   ├── test/        # Integration tests
 │   └── dist/        # Compiled binary
 ├── extension/       # Chrome extension (side panel + activity feed + CSS inspector)
-├── lib/             # Shared libraries (worktree.ts, egress-receipt.ts, context-bill.ts, redact-engine.ts, tracker-guard.ts, version-source.ts, code-intelligence/)
+├── lib/             # Shared libraries (worktree.ts, egress-receipt.ts, context-bill.ts, redact-engine.ts, tracker-guard.ts, version-source.ts, outside-voices/, code-intelligence/)
 ├── patches/         # bun `patchedDependencies` patches (playwright-core windowsHide)
 ├── docs/designs/    # Design documents
 ├── setup-deploy/    # /setup-deploy skill (one-time deploy config)

CONTRIBUTING.md

@@ -352,11 +352,11 @@ worth the review overhead.
 
 ## Multi-host development
 
-gstack generates SKILL.md files for 8 hosts from one set of `.tmpl` templates.
+gstack generates SKILL.md files for 10 hosts from one set of `.tmpl` templates.
 Each host is a typed config in `hosts/*.ts`. The generator reads these configs
 to produce host-appropriate output (different frontmatter, paths, tool names).
 
-**Supported hosts:** Claude (primary), Codex, Factory, Kiro, OpenCode, Slate, Cursor, OpenClaw.
+**Supported hosts:** Claude (primary), Codex, Factory, Kiro, OpenCode, Slate, Cursor, OpenClaw, Hermes, GBrain.
 
 ### Generating for all hosts
 
@@ -365,7 +365,7 @@ to produce host-appropriate output (different frontmatter, paths, tool names).
 bun run gen:skill-docs                    # Claude (default)
 bun run gen:skill-docs --host codex       # Codex
 bun run gen:skill-docs --host opencode    # OpenCode
-bun run gen:skill-docs --host all         # All 8 hosts
+bun run gen:skill-docs --host all         # All 10 hosts
 
 # Or use build, which does all hosts + compiles binaries
 bun run build

README.md

@@ -192,10 +192,10 @@ Each skill feeds into the next. `/office-hours` writes a design doc that `/plan-
 | `/office-hours` | **YC Office Hours** | Start here. Six forcing questions that reframe your product before you write code. Pushes back on your framing, challenges premises, generates implementation alternatives. Design doc feeds into every downstream skill. |
 | `/plan-deliverables` | **Planner** | Author per-milestone acceptance criteria, each paired with a validating test (the deliverable), and bake them into the plan. |
 | `/autobuilder-loop` | **Build Loop** | Drive an already-approved plan, spec, or backlog to completion unattended — model-routed subagents per milestone, review gates, Docker verification. |
-| `/plan-ceo-review` | **CEO / Founder** | Rethink the problem. Find the 10-star product hiding inside the request. Four modes: Expansion, Selective Expansion, Hold Scope, Reduction. |
-| `/plan-eng-review` | **Eng Manager** | Lock in architecture, data flow, diagrams, edge cases, and tests. Forces hidden assumptions into the open. |
+| `/plan-ceo-review` | **CEO / Founder** | Rethink the problem. Find the 10-star product hiding inside the request. Four modes: Expansion, Selective Expansion, Hold Scope, Reduction. Ends with the outside-voices advisory panel: independent second opinions from other AI vendors, tallied into one recommendation that never blocks. |
+| `/plan-eng-review` | **Eng Manager** | Lock in architecture, data flow, diagrams, edge cases, and tests. Forces hidden assumptions into the open. Ends with the outside-voices advisory panel. |
 | `/plan-design-review` | **Senior Designer** | Rates each design dimension 0-10, explains what a 10 looks like, then edits the plan to get there. AI Slop detection. Interactive — one AskUserQuestion per design choice. |
-| `/plan-devex-review` | **Developer Experience Lead** | Interactive DX review: explores developer personas, benchmarks against competitors' TTHW, designs your magical moment, traces friction points step by step. Three modes: DX EXPANSION, DX POLISH, DX TRIAGE. 20-45 forcing questions. |
+| `/plan-devex-review` | **Developer Experience Lead** | Interactive DX review: explores developer personas, benchmarks against competitors' TTHW, designs your magical moment, traces friction points step by step. Three modes: DX EXPANSION, DX POLISH, DX TRIAGE. 20-45 forcing questions. Ends with the outside-voices advisory panel. |
 | `/design-consultation` | **Design Partner** | Build a complete design system from scratch. Researches the landscape, proposes creative risks, generates realistic product mockups. |
 | `/review` | **Staff Engineer** | Find the bugs that pass CI but blow up in production. Auto-fixes the obvious ones. Flags completeness gaps. |
 | `/investigate` | **Debugger** | Systematic root-cause debugging. Iron Law: no fixes without investigation. Traces data flow, tests hypotheses, stops after 3 failed fixes. |

TODOS.md

@@ -2281,18 +2281,6 @@ Shipped: Default model changed to Sonnet for structure tests (~30), Opus retaine
 
 Shipped as v0.5.0 on main. Includes `/plan-design-review` (report-only design audit), `/qa-design-review` (audit + fix loop), and `/design-consultation` (interactive DESIGN.md creation). `{{DESIGN_METHODOLOGY}}` resolver provides shared 80-item design audit checklist.
 
-### Design outside voices in /plan-eng-review
-
-**What:** Extend the parallel dual-voice pattern (Codex + Claude subagent) to /plan-eng-review's architecture review section.
-
-**Why:** The design beachhead (v0.11.3.0) proves cross-model consensus works for subjective reviews. Architecture reviews have similar subjectivity in tradeoff decisions.
-
-**Context:** Depends on learnings from the design beachhead. If the litmus scorecard format proves useful, adapt it for architecture dimensions (coupling, scaling, reversibility).
-
-**Effort:** S
-**Priority:** P3
-**Depends on:** Design outside voices shipped (v0.11.3.0)
-
 ### Outside voices in /qa visual regression detection
 
 **What:** Add Codex design voice to /qa for detecting visual regressions during bug-fix verification.
@@ -2786,6 +2774,12 @@ needs one paid run to validate, so it didn't ride the ship.
 
 ## Completed
 
+### ✅ DONE (fork: outside-voices panel M1, 2026-08-30): Design outside voices in /plan-eng-review
+
+**What:** Extend the parallel dual-voice pattern (Codex + Claude subagent) to /plan-eng-review's architecture review section.
+
+**Completed:** Unreleased fork milestone M1 (feat/outside-voices-panel, 2026-08-30). Superseded by the N-voice `{{OUTSIDE_VOICES:surface=eng}}` advisory panel: external CLI voices via `bin/gstack-panel` (codex default-on; grok/gemini default-off pending live write-denial canaries), plus the Fable subagent and a native Claude pass dispatched by the orchestrator, all tallied by `bin/gstack-vote`. Wired into /plan-eng-review, /plan-ceo-review, and /plan-devex-review. Advisory only; routes into the existing human gate. See docs/designs/OUTSIDE_VOICES_PANEL.md.
+
 ### ✅ DONE (v1.68.1.0): Stop-hook registration pins the setup-time absolute path
 
 **Priority:** P1 (was filed Effort S, scoped to the Stop hook — shipped as the full defect class)