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 shimo4228/claude-harness --skill herdr-delegategit clone --depth 1 https://github.com/shimo4228/claude-harnessWrote 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/shimo4228/claude-harness/herdr-delegate)<a href="https://agentmods.dev/skills/shimo4228/claude-harness/herdr-delegate"><img src="https://agentmods.dev/badge/skills/shimo4228/claude-harness/herdr-delegate/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/shimo4228/claude-harness/herdr-delegate"><img src="https://agentmods.dev/badge/skills/shimo4228/claude-harness/herdr-delegate.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 | $0.00225 | $0.02966 |
| Opus 5 | $0.00112 | $0.01483 |
| Sonnet 5 | $0.00045 | $0.00593 |
| Haiku 4.5 | $0.00022 | $0.00297 |
Grade B, and why
herdr-delegate scanned grade B with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
`--ignore-user-config --ignore-rules` を外さない。`~/.codex/config.toml` の How it starts
The opening of the file, as written. The whole thing — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Herdr Delegate — 別 CLI エージェントへのタスク委譲
Herdr の pane に Codex 等の対話セッションを立て、指示書ファイルを渡して実装を委譲し、 完了を監視して検収する。Code Sovereignty は委譲しない — 書くのは相手、 レビュー・検証・コミット判断はこちら(呼び出し側の Claude)が握る。
ゲート(両方必須)
test "${HERDR_ENV:-}" = 1— Herdr 管理下の pane にいること- ユーザーの明示指示があること(「Codex にやらせて」等)。有益そうというだけで
自発的に委譲しない(
rules/common/agents.mdの委譲ゲートと同一)
なぜ headless(codex exec)ではなく Herdr セッションか
理由は substrate が吸収していない 3 点に絞る。長時間実行それ自体は理由にならない
(Bash tool は timeout に達したコマンドを kill せず background へ移すし、上限は
BASH_MAX_TIMEOUT_MS で変えられる — 2026-08-01 に公式 tools-reference で確認)。
- 別ベンダのモデルを使うこと自体 — Claude Code 側の並列機構(subagents / Agent teams /
/batch)はすべて Claude モデルで、cross-vendor 委譲の公式面は無い。脱相関が要るなら プロセスを分けるしかない - 対話セッションの永続性と差し戻し — 公式の background 実行(
claude --bg)は プロンプトを起動時に渡し切る形で、走っている相手に後から文字を送る API が無い。 同一文脈に差し戻して直させる運用には公式の代替が無い(herdr agent promptで 同じセッションに追記できるのがこの経路の核) - 人間の可視性 — 途中経過が pane に出て、ユーザーの追加指示・中断(esc)が効く
捏造報告に注意(2026-07-31 実測): headless 実行が途中で打ち切られた際、打ち切り直前までの 文脈で実在しない成果物の完了報告を捏造する事例を確認した(「92 テスト green・ファイル 作成済み」→ 実際は tree 無変更)。これは実行機構を変えても消えない性質なので、 下の検収規律(ground truth は git)で受ける。
手順
1. 指示書をファイルに書く
プロンプトは herdr コマンドの引数に直接埋めず、scratchpad にファイルとして書く (長文の引用崩れ防止・監査記録・再実行可能性)。指示書に必ず含めるもの:
- 成果物の定義と置き場所(ファイルパスまで具体的に)
- リポジトリ固有の制約(lint が機械拒否する語彙・スタイル規約・並行性設定など。 相手はこの repo の CLAUDE.md を読むとは限らない — 必要な不変条件は指示書に転記する)
- 検証手順(build / test / lint のコマンド列)と「自分で green にしてから完了報告」
git commit はしないこと(ワーキングツリー残置。検収とコミットはこちら)- sleep 禁止・決定論などテスト規律(
rules/common/testing.mdから該当分を転記) - サンドボックスで実行できない検証の代替を指定する — Codex の seatbelt からは
iOS Simulator(CoreSimulatorService)に接続できない(2026-07-31 実測)。
simulator を要するテストは「コンパイル検証(
xcodebuild build-for-testing)までを 相手、実行は検収側」と指示書に明記し、迂回(sandbox 解除要求)をさせない
2. pane を割って agent を立て、指示書パスを渡す
herdr pane split --current --direction right --cwd "$PWD" --no-focus
# → .result.pane.pane_id を読む
herdr agent start <name> --kind codex --pane <pane-id> \
-- --ignore-user-config --ignore-rules -s workspace-write
herdr agent prompt <name> "指示書が <指示書の絶対パス> にある。読んでそのとおり実装し、検証を自分で回して green にしてから完了報告。git commit はしないこと。"
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.
- 8d ago First seen · 146 lines · 225 tokens per session scan B f2298e900032
herdr-delegate is a skill published in the GitHub repository shimo4228/claude-harness (2 stars, last pushed 2d ago), licensed MIT. It adds 225 tokens to every session and 2,966 once invoked, about $0.0011 per session on Opus 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 2026-08-31.
Other skills, from other repositories
agent-adoption-triage
Five-question triage for routing AI work to the right architecture — script, algorithmic search, LLM workflow, or autonomous agentic loop — before any agent is built. Use when someone says "let's build an agent for X", when deciding whether a use case actually needs an autonomous agent, when reviewing an agent…
job-description-skill
A job-description analysis skill that turns a vacancy listing and a candidate’s background into an HTML report about fit, gaps, interview topics, and application strategy.
offer-compare-skill
A decision assistant for comparing two or more job offers across pay, growth, company strength, team risk, promotion, AI exposure, and lifestyle. Total compensation means the combined value of salary, bonuses, shares, and sign-on payments.
offer-toolkit-skill
A set of job-search tools covering the path from finding a vacancy to accepting an offer. It includes separate tools for job searches, job-description analysis, tailored resumes, behavioral interviews, offer comparisons, and salary negotiation.
bq-skill
A coaching workflow for building a reusable library of truthful stories for behavioral job interviews. It uses structures such as STAR—Situation, Task, Action, Result—and can connect stories to a job description and résumé.
linkedin-job-search-skill
A job-search assistant that uses a sample job description and your resume to find and rank similar public LinkedIn Jobs listings. LinkedIn Jobs is LinkedIn’s section for advertised employment positions.