<< All versions
Skill v1.0.1
LLM-judged scan95/100sunrain520/spec-first/spec-team-standards-governance
+16 new
──Details
PublishedJuly 8, 2026 at 02:05 PM
Content Hashsha256:dbdbe94f95178a56...
Git SHA437bb9e4bfa4
Bump Typepatch
──Files
Files (1 file, 8.4 KB)
SKILL.md8.4 KBactive
SKILL.md · 92 lines · 8.4 KB
version: "1.0.1" name: spec-team-standards-governance description: "Govern source-backed team development standards: query confirmed rules, audit health, initialize or draft evidence-based candidates, and prepare promotion/deprecation proposals. Do not restore legacy standards workflow entrypoints, skills/spec-standards/, or treat advisory candidates as hard context."
Spec Team Standards Governance
Use this standalone skill when the user asks to query, initialize, audit, propose, promote, or deprecate source-backed team development standards. It is a source-maintenance method, not a public spec-* workflow and not the retired spec-standards workflow.
When To Use / When Not To Use
- Use for standards governance work on
docs/contracts/team-standards.md,docs/standards/**, candidate evidence, health audits, or promotion/deprecation proposals. - Do not use for ordinary code/doc review, implementation, PRD/plan authoring, or workflow execution; those route to their own public
spec-*workflow. - Do not use to create, restore, or recommend legacy standards command spellings, the retired
spec-standardsworkflow,skills/spec-standards/, or.spec-first/standards/. - Do not turn
observed,suggested,imported,conflict,confirmed-draft, replay results, or high confidence into enforceable hard context.
Hard Boundaries
- Do not create legacy standards workflow entrypoints,
skills/spec-standards/or.spec-first/standards/. - Do not edit generated runtime mirrors such as
.claude/,.codex/or.agents/skills/. - Standalone direct use defaults to report/proposal-only. Durable source mutation requires an active
spec-workor equivalent source-edit workflow, ordinary diff review,CHANGELOG.md, and focused tests. - Scripts or structured steps may collect deterministic/advisory facts; the LLM decides semantic applicability and promotion posture.
- Only
trust=confirmed,lifecycle_state=activeand scope-matched standards can become hard context.observed,suggested,imported,conflictandconfirmed-draftare not enforceable. - Confidence score is not authority. High-impact governance, conflicts and owner-unresolved rules require owner/ADR/design-note handling.
Modes
| Mode | Purpose | Output | Source mutation boundary | |
|---|---|---|---|---|
query | Return relevant standards for a workflow slice | Filtered rule IDs, matched files, excluded reasons, fallback and limitations | read-only | |
init | Initialize brownfield candidate notes from explicit sources | acquisition notes, candidate patch preview, conflicts | proposal-only; V1 writes candidates only inside source-edit workflow | |
propose | Draft candidates from repeated issues, incidents or source refs | suggested / observed candidate cards and decision trace | proposal-only | |
promote | Prepare confirmed-draft or confirmed patch proposal | promotion proposal with authority tier, gates, owner status and index patch preview | actual confirmed/index writes require source-edit workflow | |
deprecate | Prepare lifecycle downgrade | deprecated/archive patch preview and invalidation evidence | actual lifecycle/archive writes require source-edit workflow | |
audit | Check standards health | advisory drift/conflict/stale-owner/no-load-all report | no blocking gate | |
eval/replay | Evaluate an existing acquisition run with replay, retrieval, noise and owner-edit evidence | eval cases, threshold results, limitations and promotion evidence | read-only or validation artifact; never a promotion gate by itself |
Reference Loading Map
Read only the references required by the active mode. Never load every reference by default.
| Scenario | Read references | Do not default-read | |
|---|---|---|---|
query | references/meta-prompt-governance.md, references/loading-and-consumption.md; read authority-tiers.md or promotion-and-conflicts.md only if conflict/tier interpretation is needed | initialization, acquisition scoring, interview or replay details | |
init | references/meta-prompt-governance.md, references/initialization.md, references/acquisition-quality.md, references/output-risk-profile.md | replay/golden samples, full lifecycle details | |
init with V2 pilot input | references/source-matrix.md, references/role-interview-playbook.md when source mix or owner questions affect acquisition quality | unrelated surface refs, replay details unless samples exist | |
propose | references/meta-prompt-governance.md, references/acquisition-quality.md, references/adaptive-expansion.md, references/promotion-and-conflicts.md; use references/source-matrix.md when source authority is uncertain | all standards files, unrelated surface refs | |
promote | references/meta-prompt-governance.md, references/authority-tiers.md, references/promotion-and-conflicts.md, references/loading-and-consumption.md | initialization playbook, broad evidence collection | |
deprecate | references/meta-prompt-governance.md, references/lifecycle.md, references/promotion-and-conflicts.md, references/authority-tiers.md | acquisition task pack details | |
audit | references/meta-prompt-governance.md, references/loading-and-consumption.md, references/lifecycle.md, references/output-risk-profile.md; for V2 ledgers read references/source-matrix.md and references/validation-and-replay.md | role interviews unless owner gaps require follow-up | |
eval/replay | references/meta-prompt-governance.md, references/validation-and-replay.md, references/output-risk-profile.md, evals/README.md, and only the named eval cases | init/propose/promote details not needed for the eval question |
Workflow
- Parse mode from the user request. If unclear, default to
queryfor lookup requests andauditfor health-check requests; otherwise ask one short clarification. - Read
docs/contracts/team-standards.md. - For standards selection, read
docs/standards/index.mdbefore rule files. If the index is missing, stale or scope is unknown, use the fallback modes from the contract. - Read only mode-specific references from the loading map.
- Produce the mode output per the Output Contract, including the mode-specific fields for the active mode; never emit a field you cannot back with
source_refs_used. - If source edits are authorized by an outer source-edit workflow, keep writes scoped to
docs/standards/**, this skill's source files, tests, docs andCHANGELOG.md; never patch runtime mirrors.
Output Contract
This section is the single source of truth for output fields. Every output includes mode; status (completed, degraded, blocked, or proposal-only); source_refs_used; fallback_mode; limitations; and next_action.
Add the fields required by the active mode:
query/audit:matched_rule_ids,matched_files,excluded_rule_ids,uncertainty_reason.init/propose:candidate_ids,authority_tier, owner status,why_not_confirmed, pre-write gate result,decision_trace.promote/deprecate:proposal_ids,gate_results,confidence.signals,autonomy.mode,outcome,decision_trace, owner status, and the diff/source files that a source-edit workflow would update.eval/replay:acquisition_id,replay_case_ids, single extraction target, source anchors, evidence quality, replay status, owner-edit status, and whether samples were sufficient.
Failure Modes
| Reason | Response | |
|---|---|---|
contract-missing | Degrade to host instructions and direct source reads; do not treat standards files as hard context | |
index-missing | Do not scan all docs/standards/**; ask for index creation or use explicit user-requested rule refs | |
scope-uncertain | Load only shared/high-priority safe summary; report uncertainty | |
conflict-present | Return conflict refs and owner next action; do not enforce | |
prewrite-gate-failed | Do not write candidate content; return report-only limitation | |
source-edit-not-authorized | Produce patch preview only | |
not-enough-sample | Record degraded eval evidence; do not claim replay/retrieval pass | |
owner-unavailable | Keep candidates advisory or prepare owner questions; do not infer owner approval |
Non-Goals
- No CLI command, route-map entry, public workflow catalog entry or runtime mirror patch.
- No automatic rule mining to confirmed policy.
- No PR replay, retrieval eval, role interview or V2 ledger claim unless real pilot inputs exist.