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 atman-33/workhub --skill persona-crewgit clone --depth 1 https://github.com/atman-33/workhubWrote 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/atman-33/workhub/persona-crew)<a href="https://agentmods.dev/skills/atman-33/workhub/persona-crew"><img src="https://agentmods.dev/badge/skills/atman-33/workhub/persona-crew/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/atman-33/workhub/persona-crew"><img src="https://agentmods.dev/badge/skills/atman-33/workhub/persona-crew.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 | $0.00059 | $0.01167 |
| Opus 5 | $0.00030 | $0.00583 |
| Sonnet 5 | $0.00012 | $0.00233 |
| Haiku 4.5 | $0.00006 | $0.00117 |
Grade A, and why
persona-crew scanned grade A with 0 findings 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 9d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 95 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Derived from genshijin (MIT, InterfaceX-co-jp) — https://github.com/InterfaceX-co-jp/genshijin
persona-crew は圧縮形式で出力する3つの subagent プリセット。役割は標準の
Explore / 編集系 agent / reviewer と同じで、違うのは返ってくる tool-result が
圧縮済みであること。委譲のたびに主コンテキストの消費が縮む。
キャラクター非依存。 agent の出力は口調ではなく圧縮形式に従う。
使い分け
- 「X の定義はどこ / Y を呼ぶ箇所 / Z の全用法」 →
persona-investigator - 同上に加えてアーキテクチャの解説や提案も欲しい →
Explore(標準) - スコープが明確な surgical 編集、2ファイル以下 →
persona-builder - 新機能 / 3ファイル以上 / 横断リファクタ → 主スレッド、または
heavy-implementer - diff / branch / file のバグレビュー →
persona-reviewer - 根拠と代替案つきの深いコードレビュー → 標準の Code Reviewer
- 1行で答えが確定している内容 → 主スレッドで完結。委譲しない
判断基準は単純で、subagent の出力を 1/3 のトークンで欲しいなら persona-crew、 散文で欲しいなら標準の agent。
なぜ存在するか
subagent の tool-result は主コンテキストにそのまま注入される。標準の Explore が
散文で 2k トークン返せば、毎回 2k を主コンテキストが負担する。同じ発見が
persona-investigator なら 700 トークン前後で返る。1セッションで20回委譲すると、
コンテキスト枯渇か完走かの差になる。
出力契約
主スレッドが agent ごとに前提にしてよい形式。
persona-investigator:
<Header>:
- path:line — `symbol` — 短い注記
集計: <counts>。
該当なしは No match.。必ずファイルパスが先頭、行番号つき、シンボルはバッククォート。
path:\d+ で grep できる。
persona-builder:
<path:line-range> — <変更 10語以内>。
verified: <re-read OK | mismatch @ path:line>。
または終端行のいずれか: too-big. / needs-confirm. / ambiguous. / regressed.
persona-reviewer:
path:line: <emoji> <severity>: <問題>。<修正>。
totals: N🔴 N🟡 N🔵 N❓
指摘なしは No issues.。ファイル順、ファイル内は行の昇順。
チェイニング
位置特定 → 修正 → 検証(最頻)
persona-investigatorで対象箇所の一覧を取る- 主スレッドが1〜2箇所を選び、パスを
persona-builderに渡す persona-reviewerが diff を監査する
並列スカウト(調査範囲が広いとき)
1メッセージで persona-investigator を2〜3個並列起動する(定義 / 呼び出し元 / テスト
のように角度を変える)。集約は主スレッドで行う。
単発編集(箇所が既知のとき)
investigator を飛ばし、persona-builder に直接 path:line を渡す。
禁止
- ファイルを特定せずに
persona-builderを使わない。先に investigator を通す。 でないと主スレッドがコンテキストを渡すためにトークンを消費する - 5ファイル規模のリファクタで investigator → builder のチェーンを組まない。
builder は
too-big.を返してターンを浪費する persona-reviewerに「全般的なフィードバック」を求めない。findings しか返らない- 散文を期待しない。persona-crew の出力は構造化されており、時に読みにくい。 人間がそのまま読む場合は主スレッドが言い換える
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.
- 9d ago First seen · 95 lines · 59 tokens per session scan A db484ce65648
persona-crew is a skill published in the GitHub repository atman-33/workhub (2 stars, last pushed today), licensed MIT. It adds 59 tokens to every session and 1,167 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
gonavi-cli
Operate databases through the GoNavi headless CLI — the gonavi executable shipped in verified GitHub Release archives. Covers listing/adding/importing saved connections, running SQL queries against saved connections or ad-hoc connection files, exporting result sets to csv/json/md/html/xlsx, batch-executing SQL files…
xiaohongshu-image-creator
An image-making assistant for Xiaohongshu, a Chinese social platform for lifestyle, product, and educational posts. It creates vertical covers and supporting images matched to the post’s topic, audience, and visual style.
x-cr
A software-correctness investigation skill for finding why code behaves differently from what was expected. It uses evidence from code paths, specifications, tests, logs, and changes to assess possible causes.
design
Create a doc-as-code design package from a PRD or SPEC. Conditionally generates C4 diagrams (Context/Container/Component), sequence diagrams, ER diagram + Data Dictionary, OpenAPI 3.0, AsyncAPI 3.0, ADRs, domain glossary, state diagrams, and deployment view as Mermaid-rendered Markdown files. Use when PM mentions…
recall
Retrieve session memory across 4 modes — TEMPORAL (date browsing), KEYWORD (FTS lookup), DEEP (cross-session synthesis), LOAD (specific file). Trigger on "what did I do yesterday/last week", "why did we...", "remind me...", "find the memo about...", "last time we...", "patterns across projects", "load the X topic". Do…
shipyard-handoff
Captures session context into .shipyard/HANDOFF.md so the next session can resume without losing progress.