Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add thkt/dotclaude --skill implementgit clone --depth 1 https://github.com/thkt/dotclaudeWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/thkt/dotclaude/implement)<a href="https://agentmods.dev/skills/thkt/dotclaude/implement"><img src="https://agentmods.dev/badge/skills/thkt/dotclaude/implement/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/thkt/dotclaude/implement"><img src="https://agentmods.dev/badge/skills/thkt/dotclaude/implement.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00068 | $0.03181 |
| Opus 5.5 | $0.00027 | $0.01272 |
| Sonnet 5.5 | $0.00014 | $0.00636 |
| Haiku 4.5 | $0.00007 | $0.00318 |
Grade B, and why
implement scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 16d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
4. sandbox の設定を確認する。check や撮影はローカルの web server の socket を開くことがあり、sandbox がその bind を拒むと check は毎回失敗し、回数上限の無い修正ループが止まらない。存在する設定ファイルを `~/.claude/settings.json`、対象 repo の `.claude/settings.json`、`.claude/settings.local.json` How it starts
The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/implement - 合意済み Issue から PR 作成
/scoping が合意させた Issue を implement workflow に渡し、実装から draft PR と CI 確認までを決定論的に進める。このスキルは起動の前の照合と、workflow が外に残す公開本文の照合・ready 切替を担う。
$ARGUMENTS は Issue 番号か URL。--no-publish が付いたら公開を切って起動し、workflow は verified_local で止まる。対象 Issue を特定できないときは AskUserQuestion で確認する。対象 repo は現在のリポジトリ。要求の変更が要ると分かったら止め、/scoping へ戻す。合意済みの範囲内の作業ごとに許可を聞き直さない。
確認できる事実は調査し、ユーザーにしか分からない事実や未確定の意図は早めに質問する。指示文を理由に確認や停止が必要な場合は、実際に読んだ文書のパスと該当文を示し、明示された条件と自分の解釈を区別する。
Phase 1: 起動前の照合
gh issue view <ref> --json number,title,body,state,urlで Issue を取る。state が OPEN で title と body が空でないことを確認する。git rev-parse --show-toplevelで対象 repo の絶対パスを決める。.dotagents.jsonが commit 済みであることを確認する。workflow は repository・remote・baseBranch・setup・check・ciChecks・capture を持つ Codex の対象契約で検査する。無ければ package.json の scripts や CI 定義から内容を仮説付きで AskUserQuestion に出し、合意した内容をユーザーの判断で commit してから進める。未 commit の差分があれば stash や移植をせず、扱いを尋ねる。Issue の必要媒体を${CLAUDE_SKILL_DIR}/references/capture.mdの表でcaptureの設定と照合し、食い違えば起動せず/scopingへ戻す。- sandbox の設定を確認する。check や撮影はローカルの web server の socket を開くことがあり、sandbox がその bind を拒むと check は毎回失敗し、回数上限の無い修正ループが止まらない。存在する設定ファイルを
~/.claude/settings.json、対象 repo の.claude/settings.json、.claude/settings.local.jsonの順に並べる。それらをjq -s 'reduce .[] as $s ({}; . * $s) | .sandbox | {enabled, allowLocalBinding: .network.allowLocalBinding}'に渡す。enabledが true でallowLocalBindingが true でなければ起動せず、sandbox.network.allowLocalBinding: trueを設定するよう伝えて止める。この設定は再起動なしで効く。 - 固定する参照を選ぶ。
${CLAUDE_SKILL_DIR}/../../rules/conventions/DOCUMENTS.md§ Read and retain に従い、関連する wiki と判断記録を選ぶ。wiki は${CLAUDE_SKILL_DIR}/../scribe/scripts/find_wiki_rule.ts docs/wiki <Issue の語> <触れそうなパス> --scene implementのmatchedとscenesのページを読む。選んだページと判断記録を、Issue と現行コードに照合する。${CLAUDE_SKILL_DIR}/../scoping/references/session.md§ 調査成果の引き継ぎに従い、Issue が参照する調査報告も加える。参照に使えるのはdocs/research/・docs/wiki/・docs/decisions/配下の Markdown だけ。各参照の blob をgit rev-parse HEAD:<path>で、開始 commit をgit rev-parse HEADで記録する。必要な報告が HEAD に commit されていなければ、引き継ぎ未完了として止める。 - 既存 PR を直す依頼なら、この Phase を離れて § 既存 PR の修正へ回す。
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 16d ago Changed 3d9eecb2737b
- 17d ago Changed 2d05a3bced70
- 18d ago First seen · 75 lines · 68 tokens per session scan B 64bf3519dfc1
implement is a skill published in the GitHub repository thkt/dotclaude (12 stars, last pushed 17d ago), licensed MIT. It adds 68 tokens to every session and 3,181 once invoked, about $0.0003 per session on Opus 5.5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-28.
Other skills, from other repositories
merge-ready
Run all quality gates before merge — git hygiene, documentation completeness, code review, security audit, build, E2E, goal-backward verification, doc accuracy and UI/UX — then write the changelog entry.
develop-feature
Autonomously implement a complete feature from request to merge-ready — documentation phase, wave-based TDD implementation, then all quality gates, without stopping for human input.
implement-slice
Implement the next smallest slice from the current plan using TDD — tests first, then code, then verification, then an atomic commit. Reads the plan from the scratchpad.
implementation-review
Pre-commit quality gate. Invoke before every git commit, after /flagrare:staleness-audit. Seven checks, plan gaps, use-case coverage gaps, missing test scenarios, test philosophy violations (Kent Dodds Testing Trophy), SOLID violations, Clean Code violations, and security vulnerabilities. Each check is delegated to a…
staleness-audit
Pre-commit audit. Invoke before every git commit. Diffs the staged change against the repo's "canaries" (README, decision log, roadmap, public API, doc comments, release notes, test discipline) and surfaces anything that's drifted out of sync. Updates the obvious drifts in place; flags judgement calls for the operator.
wrap-up
Post-implementation quality gate. Runs automated checks (tests, lint, types), invokes /flagrare:implementation-review for the seven-check parallel review, culls every code comment in the diff that is not a trap-preventer, then performs additional SOLID and Clean Code review on any findings not covered. Use when the…