Two changes:
1. CONFIDENCE_THRESHOLD 0.8 → 0.6 — catches borderline recommendations
that previously slipped through. Driver: brain-retro #10 shows 0%
single-node-skill follow-through, suggesting hook needs to fire more.
2. Inline escape hatch — 'router-skip: <reason 50+ chars>' in assistant text.
Per-tool scope (does not affect other tools in same turn). Replaces
the documented 'override: <reason>' hint which was a self-bypass
loophole — high-friction 50+ char justification discourages reflexive use.
Per Level 2 of plan docs/superpowers/plans/2026-05-28-router-discipline-level-1-2.md.
Legacy tests flipped (2 tests):
- 'allows when confidence exactly 0.7 (raised threshold)' →
'BLOCKS when confidence exactly 0.7 (above new threshold 0.6)'
- 'allows when confidence 0.75 (still under raised threshold)' →
'BLOCKS when confidence 0.75 (above new threshold 0.6)'
These tests previously asserted block:false at 0.7/0.75 under the old 0.8
threshold; with 0.6 threshold they now correctly assert block:true.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three brain-governance hardening changes from retro #8 follow-up:
1. enforce-classifier-match: confidence threshold raised 0.7→0.8 (was producing false-positives on borderline LLM recommendations like #3 GitHub MCP for local debug, #36 adr-kit for status readouts). 2 new vitest tests cover boundary values 0.7 and 0.75 (now allowed).
2. enforce-chain-recommendation (NEW): PreToolUse hook blocking mutating tool calls when router gave recommended_chain length >= 2 and controller is not expanding it. Allows pass when: any chain node already invoked, inline 'chain-override: <reason>' present, or global override-phrase in user prompt. 20 vitest tests cover empty chain, single-node bypass, override variants, alias resolution, mixed numeric/string ids.
3. registry-load.test.mjs: bump expected counts 85→86 nodes / 77→78 active (collateral fix after parallel session added #86 graphifyy in 27289c05).
Full vitest tools-sweep: 1022/1022 GREEN.
Reviewer APPROVE on spec compliance + code quality (non-blocking observations: test count mis-report in implementer's claim 33→20 actual, hardcoded 'superpowers:' alias prefix, no direct test for extractCalledSkillIds — deferred).
Hook activation in .claude/settings.json deferred — controller will register separately based on owner's choice (block / warn-only / defer).
Brain-retro #5 candidate C, hole 4: enforce-classifier-match.mjs main()
read only state.classification.recommended_node, which is null for
prefilter/regex classifier sources. When triggers_matched[0] contained a
recommendation, the rule was bypassed.
Added fallback: if recommended_node is null, use triggers_matched[0]. decide()
already accepts null confidence on this path (only numeric < 0.7 blocks).
Brain-retro #5 candidate C, hole 2: enforce-classifier-match.mjs's
MUTATING_TOOLS set missed Task/Agent, so delegating mutations via Task()
bypassed the rule. Added Task and Agent to the set; nodeMatches already
handles Task.subagent_type matching.
Regression test asserts Task with matching subagent_type does NOT block
(keeps the existing nodeMatches Task path intact).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Brain-retro #5 candidate C, hole 1: enforce-classifier-match.mjs allowed
the agent to bypass the rule by writing 'override: <reason>' in its own
response (self-override = no enforcement). The user-vocabulary override
phrases in enforce-override-vocab.json remain the only legitimate path.
Added regression test asserting block on assistantText override when user
prompt has no override phrase.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Дмитрий