feat(voyage): define high-effort behavior + amend brief Non-Goal/SC1 + coordinator normalization (Decision B)
Wave 6 / Step 10 — autonomy-gated. Operator confirmed: gemini-bridge substitution for plan-critic doubling AND SC1 amendment to resolver-invariant encoding (decisions.local.json recorded). - commands/trekplan.md: gemini-bridge plan-review Pass 2 on post-revision plan in high-effort mode (replaces fragile plan-critic doubling per risk-assessor). - commands/trekresearch.md: full swarm + contrarian-researcher + gemini-bridge always-on. - commands/trekreview.md: skip Pass 3 reasonableness + invoke coordinator normalization rule. - commands/trekexecute.md: gates_mode = closed (strict manifest-audit, main-merge pauses); flag override still wins. - agents/review-coordinator.md: Pass 3 high-effort normalization — substitute unknown rule_key with PLAN_EXECUTE_DRIFT, preserve original in original_rule_key. - .claude/projects/2026-05-13-trekflow-solo-lane/brief.md (gitignored, not committed): Non-Goal amendment locks low/high tiers; SC1 amendment authorizes resolver-invariant interpretation. - tests/lib/doc-consistency.test.mjs: +4 pins for the "### High-effort behavior (v5.1.1)" heading per command. Tests: 578 pass, 0 fail, 2 skipped (+4 from 574). Closes #7 (operator-gated decisions captured + coordinator normalization landed). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
This commit is contained in:
parent
07ae1e30e9
commit
c1b7bad389
6 changed files with 114 additions and 0 deletions
|
|
@ -112,6 +112,18 @@ Drop findings that fail ANY of these tests:
|
|||
In `quick` mode, skip this pass entirely. Note the skip in the
|
||||
Executive Summary so the reader knows reasonableness was not applied.
|
||||
|
||||
**High-effort normalization (v5.1.1):** When the review is invoked
|
||||
under high-effort mode (`phase_signals[review].effort: high`), Pass 3
|
||||
reasonableness filtering is bypassed. To prevent unknown rule_keys
|
||||
from polluting downstream plans, the coordinator MUST substitute any
|
||||
rule_key not exported from `lib/review/rule-catalogue.mjs:RULE_KEYS`
|
||||
with the literal string `PLAN_EXECUTE_DRIFT` (the most general drift
|
||||
category from the 12-entry catalogue). The original rule_key is
|
||||
preserved in the finding's `original_rule_key` field for diagnostic
|
||||
purposes. This normalization happens BEFORE writing review.md,
|
||||
ensuring all `rule_key` values in the final review match the
|
||||
catalogue.
|
||||
|
||||
### Pass 4 — Compute verdict
|
||||
|
||||
Count findings by severity AFTER dedup and filtering. Verdict thresholds:
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue