akc-cycle

Registry Snapshot

Repository

https://github.com/shimo4228/akc-cycle

Homepage

https://github.com/shimo4228/akc-cycle

Canonical profile

https://hol.org/registry/plugins/shimo4228%2Fakc-cycle

Publisher verification

No

Marketplace source

Unknown

Scanner

Broker fallback

Safety label

caution

Digest verified

Yes

Marketplace Source

Repo URL

https://github.com/shimo4228/akc-cycle

Marketplace path

Unknown

Source path

.

Install policy

Unspecified

npx skills add shimo4228/akc-cycle

• adr-writer

  skills/adr-writer/SKILL.md

  Record a design decision as an Architecture Decision Record (ADR) in the project's `docs/adr/` directory. Use this skill whenever the user says "let's ADR this", "record this decision", "write an ADR for X", or when context-sync Phase 3 needs to extract a buried decision. The skill resolves the target ADR directory from cwd, picks the next sequence number with no collision, writes the 7-section body (incl. `Review-when` expiry conditions) in the main loop, runs the mechanical evidence and the adr-reviewer, and updates the ADR index. Also the canonical place for the filing bar — a new ADR is written only for a decision that is hard to reverse, surprising without context and the result of a real trade-off; a change to a prior ADR that misses the bar gets a dated Note on that ADR, and everything else goes into the commit body. Works across any repo — auto-detects or creates `docs/adr/` from the repo root.

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• agent-stocktake

  skills/agent-stocktake/SKILL.md

  Audit ~/.claude/agents/*.md (subagent definitions) for description-layer residency cost, body-layer quality, suppression instructions, staleness, and substrate absorption, assigning Keep/Improve/Update/Merge/Demote-to-skill/Dissolve/Retire verdicts. Use when the user says \"audit my agents\", \"agent stocktake\", \"which agents should I retire or merge\", \"take stock of my agents\", \"review my agent definitions\", or when the model generation changed and agent bodies written for the previous one may suppress or over-constrain the current one. NOT for — skill quality → skill-stocktake; rules → rules-stocktake; cross-checking against the runtime layer → generation-audit.

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• author-calibrated-eval

  skills/author-calibrated-eval/SKILL.md

  How to build and run a loop that polishes LLM-written reading material (body text, summaries, explanatory prose) by treating the author's reading as ground truth. Fix the materials, build a hard-case set, run blind side-by-side reads by the author, measure the ceiling with a strong model's reference draft, and keep the LLM judge dedicated to a faithfulness cutoff. Invoke with /author-calibrated-eval. NOT for — designing the judge's prompt or verdict format (→ llm-as-judge), validity of measured values and thresholds (→ measurement-discipline), rewriting a README (→ readme-writer).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• context-sync

  skills/context-sync/SKILL.md

  Audit and fix project documentation — detect role overlaps between context files (CLAUDE.md, ADR, README, graph.jsonld), migrate misplaced content, check freshness against code, and create missing docs; README findings are flagged, not edited.

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• generation-audit

  skills/generation-audit/SKILL.md

  Single entry point for auditing the harness when a new Claude model takes one of its roles (judge tier / build tier / everyday session). Runs two checks in one pass — the runtime-layer cross-check (live system prompt + tool descriptions vs rules / CLAUDE.md / output style / skills / agents: conflict and redundancy) and the dated-pattern scan of every self-authored asset via `/claude-api prompt-audit` — then applies the line fixes under this harness's writing rules and hands asset-level verdicts to rules-stocktake / skill-stocktake / agent-stocktake. Invoke with /generation-audit when a new Claude model takes a harness role. NOT for — rewriting one skill (skill-creator); routine stocktakes (call them directly); rule compliance (skill-comply).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• harness-boundary

  skills/harness-boundary/SKILL.md

  A design lens for adding, changing, or reviewing a mechanism (rule / skill / hook / agent / workflow / runtime extension / prompt chain) in an agent environment: it asks which of 6 layers (model capability / skill = procedural memory / values & policy / eval / data & memory / runtime) the mechanism belongs to, why it cannot be left to the model itself, whether the next model generation will make it unnecessary, and whether it is worth keeping when the runtime is swapped (Claude Code → Pi → Codex), and returns Keep / Move / Simplify / Make temporary / Delete / Defer. Use when — \"should this go in the harness?\", \"which layer does this belong in?\", \"can't the model handle this itself?\", \"will this survive a runtime change?\", \"the harness is bloated\", \"can I add this hook / rule / workflow?\", when an implementation plan judges a task to change the harness itself (rules / skills / hooks / agents / settings in ~/.claude), or /harness-boundary. Delete / Simplify count as success. NOT for — a standalone build-or-not decision on something not yet built (→ a build-or-not judge agent), periodic audits of installed assets and Retire / Dissolve verdicts (→ rules-stocktake / skill-stocktake / agent-stocktake; this skill only passes evidence), bulk cross-checks at a model generation change (→ generation-audit), validity of loop structure (→ loop-design-check).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• jev-judgment-design

  skills/jev-judgment-design/SKILL.md

  >

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• learn-eval

  skills/learn-eval/SKILL.md

  Extract a reusable pattern from the current session, judge it Save / Improve then Save / Absorb / Drop against a grounding checklist, and route every Save to a destination something actually reaches — absorbed into an existing skill / rule / doc section, or promoted to a real skill via skill-creator. Use when the user says \"save what we learned this time\", \"run learn-eval\" or /learn-eval. There is no notes parking lot: if nothing would route to it, the verdict is Drop. NOT for mining past sessions in bulk, auditing skills (skill-stocktake), or distilling rules (rules-distill).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• llm-as-judge

  skills/llm-as-judge/SKILL.md

  Design pattern for LLM-as-judge evaluators — binary checks as evidence, one named holistic verdict, no score aggregation. Use when designing or reviewing any LLM-based quality gate, evaluator, judge prompt, or verdict schema; when a judge's rubric scores fluctuate between runs; when you catch yourself asking an LLM for a 1-5 score, averaging check results, or thresholding a satisfaction ratio. NOT for choosing whether a task needs deterministic or semantic processing, and NOT for the architecture-level judge+enforce state-mutation split.

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• measurement-discipline

  skills/measurement-discipline/SKILL.md

  Discipline for designing or evaluating measurement-based claims, thresholds, guards, experiment results and observation periods. Use when the user says 「この実験結果で判断していい？」 / "can I decide on this experiment result?", 「閾値を決めたい」 / "I want to set a threshold", 「ガード/検査を足したい」 / "I want to add a guard or check", 「1 回通ったから大丈夫」 / "it passed once, so it's fine", 「観察期間はどれくらい」 / "how long should the observation period be?", 「いつゲートを開く」 / "when do we open the gate?", 「shadow のまま何週待つ」 / "how many weeks do we wait in shadow?", 「この RFC 塩漬けでは」 / "isn't this RFC just sitting idle?", 「本番より良い」 / "it's better than production", 「本番と比べて」 / "compare it against production", when a design places a numeric threshold, a suspicion flag, or a wait-for-N-observations condition, when a candidate is compared against production, or when a claim rests on measured data. NOT for — designing the instruments themselves (read-only distributions and readings) — out of scope here; designing an LLM judge (llm-as-judge); whether a loop's structure is sound (out of scope here).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• repo-asset-stocktake

  skills/repo-asset-stocktake/SKILL.md

  Audit a project repo's non-code assets — tool configs, CI/GitHub workflows, runbooks, other docs — for assets whose consumer has vanished, and assign Keep/Update/Retire/Merge verdicts. Use when the user says \"audit my repo assets\", \"repo asset stocktake\", \"which configs/workflows/runbooks are dead\", or \"take stock of my non-code assets\".

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• review-to-lint

  skills/review-to-lint/SKILL.md

  Move the mechanically decidable items of an existing reviewer's checklist, or of past reviewers' findings, into a deterministic evidence script, and thin the reviewer down to semantic checks. Use via /review-to-lint when the author says \"turn this reviewer into a lint\" or \"find what in the history can be linted\", or when a reviewer's findings repeat mechanical items.

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• rules-distill

  skills/rules-distill/SKILL.md

  Scan installed skills to extract principles that belong in the always-loaded rules layer (environment-specific facts, wiring, and traps — not general principles the substrate already applies) and distill them into rules — append to, revise, or create rule files. Use when the user says "distill rules", "/rules-distill", "promote patterns to rules", "what principles should become rules", after installing new skills, or when a skill-stocktake surfaces recurring patterns. NOT for auditing skill quality (that is skill-stocktake) and NOT for editing a single skill (that is skill-creator).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• rules-stocktake

  skills/rules-stocktake/SKILL.md

  Audit ~/.claude/rules (always-loaded behavioral rules) for residency cost, staleness, redundancy, broken skill pointers and substrate absorption, and assign Keep/Improve/Update/Merge/Demote-to-skill/Dissolve/Retire verdicts. Use when the user says \"audit my rules\", \"rules stocktake\", \"my rules have bloated\", or when a new model generation may have made rules written for the previous one net-negative.

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• search-first

  skills/search-first/SKILL.md

  Look outside before deciding: search the live web, registries and primary sources at decision time and bring back a report the caller picks from. Use for any question that may have an answer outside this repo — choosing a library or tool (\"which one is mainstream now\"), prior implementations of a design problem (\"is there already a skill / repo that solved this problem\"), whether a paper or post's claim applies here (\"does this claim apply to us\"), what an official spec or CLI does now (\"what does the official source say\"), the current state of practice (\"where do things stand now\"), or checking a claim against its primary source (\"does the primary source really say that\"). Also on 'search first', 'is there a package for', or /search-first. Other entry points: this repo or the local machine → Explore agent; a whole article's fact-check → the writing repo's fact-checker; bug fixes, refactors and config edits → your implementation workflow (implementation-chain in the author's harness).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• skill-comply

  skills/skill-comply/SKILL.md

  Measure whether skills, rules, and agent definitions are actually followed — generates scenarios at 3 prompt strictness levels, runs agents, classifies their tool calls, and reports compliance rates with tool call timelines. Use when a rule's or skill's real adherence is in question or right after adding one — \"is this rule actually being followed?\", \"measure how often this skill fires\", \"/skill-comply\".

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• skill-creator

  skills/skill-creator/SKILL.md

  Entry point and draft gate for writing a new skill / agent definition or substantially revising one. Use when the author says 'make a skill' (「skill 作って」), 'turn this procedure into a skill' (「この手順を skill にして」), 'write an agent definition' (「agent 定義を書いて」) or 'rewrite this skill' (「この skill を書き直して」); when learn-eval decides to Promote; when a creation-notice hook, if your harness has one, detects a new skill; or via /skill-creator. Fixes the intent as one packet, draws the boundary against adjacent skills across the whole library, writes for the strongest model tier you run, has a fresh-context subagent return a non-aggregated named verdict (Publishable / Fix / Drop), and closes with the author's read-through. NOT for — auditing the existing skill set (→ skill-stocktake), measuring compliance (→ skill-comply), checking broken references and ownership (→ skill-health), syncing to a public repo (→ your publish/sync step), extracting from a conversation and deciding Save/Drop (→ learn-eval).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• skill-health

  skills/skill-health/SKILL.md

  Scan the skill library for structural debt and assemble a per-skill health table. The deterministic core finds dangling references — a SKILL.md naming a script, bash file, agent, or sibling skill that does not exist on disk — and external skills whose directory is a symlink out of the skills root, where a local fix would be overwritten by the owning package manager. It then federates the other health signals (usage and residency-fold candidates, SKILL.md size, the latest security results, missing validators) into a `skill × dimension` report and persists the scan to results.json. Use when the user says "scan skills for debt", "check for dangling/broken references in my skills", "skill health check", "do referenced scripts/agents still exist", "which skills are not really mine to edit", or "/skill-health".

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• skill-stocktake

  skills/skill-stocktake/SKILL.md

  Audit installed Claude skills for quality and surface Keep/Improve/Update/Retire/Merge verdicts. Use when the user says "audit my skills", "stocktake", "review my skills", "which skills should I retire or merge", "do a quality pass over my skills", or "/skill-stocktake". NOT for creating or improving a single skill (that is skill-creator).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md

• verify-bootstrap

  skills/verify-bootstrap/SKILL.md

  Set up machine gates in a repo (format / lint / type check / security / dependency audit / test), or audit whether existing gates have gone stale. Use when starting a new project, when a repo has no automated quality gate, when the user says \"add linting\" (「lint を入れて」), \"put a gate on this repo\" (「この repo にゲートを立てて」), \"I want type checking\" (「型チェックを入れたい」), \"set up static analysis\" (「静的解析を整備して」), \"the tools are outdated\" (「ツールが古い」), \"audit verify\" (「verify を棚卸しして」), \"set up linting\", \"add a quality gate\", \"bootstrap the toolchain\", or invokes /verify-bootstrap. The skill carries no tool table, so the same procedure works on an unfamiliar stack. NOT for — running an existing gate once (run the repo's verify entrypoint directly), or semantic code review (that belongs to the review step of your implementation workflow).

  PreviewCopy SKILL.mdDownloadRaw SKILL.md