100%

Spec / rule / skill-body sync

Goal

Propagate the new forensic-mode contract into the authored .md surfaces so the behavior is documented where consumers read it: the review-plan + review-section skill bodies, state-discipline / plan-read-discipline rules (the committed .review-plan/ carve-out + zeroed=under-scrutiny semantics), and the /zero-plan body (zeroing arms forensic scrutiny + resets forensic_reviewed). SRP/SSOT: cite the runtime SSOT, do not inline algorithm prose (script-first.md §1).

Implementation Sketch

Authored-.md hygiene applies (CLAUDE.md §NO PROSE; skill-vocabulary.md §3): prescriptive bullets/tables only, no rationale prose. The committed .review-plan/ folder needs a documented carve-out in state-discipline.md §2 (it is the one committed plan-dir working folder). The zeroed-flag semantics shift from ‘awareness-only’ to ‘under extreme scrutiny -> most review + hard-block’.

Spec References

.claude/skills/review-plan/SKILL.md; .claude/skills/review-section/SKILL.md; .claude/rules/state-discipline.md §2; .claude/rules/plan-read-discipline.md §2.1; .claude/skills/zero-plan/ (the /zero-plan command body); .claude/rules/skill-vocabulary.md (zeroed glossary if needed).

Work Items

  • review-plan/SKILL.md: document the zeroed->forensic mode (entry detection, hardship dossier, all-section fan-out, restructure-apply, the forensic_reviewed clear), citing the runtime SSOT (no inlined algorithm).
  • review-section/SKILL.md: document the forensic-seeded per-section review (the dossier seed) + that it is fired across all sections of a zeroed plan.
  • state-discipline.md §2 + plan-read-discipline.md §2.1: add the committed .review-plan/ audit-artifact carve-out (derived analysis only; not a state store) + the zeroed=under-scrutiny semantics.
  • /zero-plan command body: zeroing arms forensic scrutiny (resets forensic_reviewed=false) and the next /review-plan runs forensic mode; update the ‘double duty’ description to a triple duty (linearity + verify-first + forensic-scrutiny).
  • Run prose-lint + plan_corpus check clean on every touched .md; no rationale-prose / effort-estimate leaks.

Fresh intel (regenerated)

{ “schema_version”: 1, “target”: { “kind”: “plan-section”, “ref”: “zeroed-plan-forensic-review/s-6019a033” }, “generated_at”: “2026-06-20T04:28:24.537442+00:00”, “graph_state”: { “head_sha”: “bd1ba13c”, “last_code_import_at”: “2026-06-20T04:24:44.366Z”, “embedding_stale”: false, “insights_stale”: false, “cpg_stale”: false }, “surfaces”: {}, “summary”: “intel-package[plan-section:zeroed-plan-forensic-review/s-6019a033] agent-authored dossier”, “degraded”: false, “dossier”: { “objective_symbols”: [], “tiers”: {}, “agent_authored”: true, “difficulty”: { “class”: “routine”, “research_online”: false, “research_mode”: “auto”, “signals”: [], “plan_dir”: “/home/eric/projects/ori_lang/plans/zeroed-plan-forensic-review” } } }

(full dossier: spec-rule-skill-body-sync—s-6019a033.intel.json)