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 KerberosClaw/kc_ai_skills --skill prep-repogit clone --depth 1 https://github.com/KerberosClaw/kc_ai_skillsWrote 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/kerberosclaw/kc_ai_skills/prep-repo)<a href="https://agentmods.dev/skills/kerberosclaw/kc_ai_skills/prep-repo"><img src="https://agentmods.dev/badge/skills/kerberosclaw/kc_ai_skills/prep-repo/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/kerberosclaw/kc_ai_skills/prep-repo"><img src="https://agentmods.dev/badge/skills/kerberosclaw/kc_ai_skills/prep-repo.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00064 | $0.03506 |
| Opus 5.5 | $0.00026 | $0.01402 |
| Sonnet 5.5 | $0.00013 | $0.00701 |
| Haiku 4.5 | $0.00006 | $0.00351 |
Grade A, and why
prep-repo 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 1mo 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 — 107 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Prep Repo
English summary: Prepare releases with scope-appropriate checks. Technical documents default to Traditional Chinese with an English summary for GitHub; explicit language requirements are acceptance criteria, while separate bilingual READMEs retain their languages.
先判斷這次要交付什麼，再檢查它能否安全、正確地公開。真正的機敏外洩、不可執行的安裝指引、錯誤的功能宣稱需要修；未約定的格式偏好不應讓已可交付的工作無限延長；使用者明確指定的語系／摘要格式屬於驗收要求，不能降級成可略過的修飾。
範圍與分工
- 從使用者要求、repo 規範與實際檔案確認目標、visibility、交付物及已有授權。已講清楚的決定不重問；資訊確實不足才釐清。
- 先查 working tree、branch、remote、現行 README／安裝方式與發布流程，保留無關的 dirty changes。只修改此次授權的 repo／檔案；多 repo 任務各自核對範圍。
- 私庫與公開匯出分開處理；不要因為準備 OSS 就把私庫改 public、複製整條 private history，或把私人操作手冊當公開文件。
- 既有系統缺技術文件可用
project-docs（doc-qaplugin）；新功能設計走spec，根因不明的 bug 走diagnose。不要把簡單文件修正升級成整套工程流程。
先分類，再修正
| 結果 | 判準 | 處理 |
|---|---|---|
| 阻擋發布 | 具體機敏外洩、預定交付物無法安裝／使用、必要驗證失敗、重要功能或安全宣稱失實 | 提供證據、影響和最小修法，修正並驗證後才發布 |
| 待確認／未驗證 | 疑似敏感值無法判定、缺少必要權限或環境、工具失敗／跳過 | 說明缺口；必要檢查未完成不能寫成通過 |
| 一般改善 | 無功能影響的命名、badge、措辭或排版偏好 | 按範圍和成本調整，不當作新 gate |
| 不適用 | 此交付物沒有該能力或契約 | 註明理由，不為填 checklist 新增能力 |
沿用使用者／repo 的嚴重度定義；沒有定義時不要把所有 blocker 都叫 P0。第一輪合併回報已知問題；修正後重驗受影響項與整體發布檔案邊界，不重跑沒有新理由的全部檢查，也不每輪追加無關修飾。遵守使用者指定的修正輪次與停止條件；沒有指定時，以能否提出有證據、不同於已失敗嘗試的最小修法決定是否繼續。反覆無進展就回報殘餘問題，不宣稱收斂。
依交付物挑檢查
| 檢查面 | 要確認的事 | 不要機械套用的要求 |
|---|---|---|
| README | 用途、成熟度、必要依賴、可重現的 quick start、限制與文件入口符合現況；重要目錄樹與實際檔案相符 | 標題可以是專案名稱；趣味標題、固定位置 badge 不是必要條件 |
| 語言與導覽 | 技術文件預設正體中文（臺灣用語）；發布到 GitHub 的文件開頭放簡短英文摘要，正文用正體中文。私庫去敏匯出也沿用此規則。雙語 README 維持英文版／中文版內容同步與互連，本 repo 使用 README.md／README_zh.md 與 正體中文／English 連結 |
獨立英文 README、程式碼、指令、API／schema 識別字與授權原文保留原樣。使用者／組織明確指定其他語系才覆蓋預設；現有文件是英文不算例外。核對本次交付，不為小修翻譯未授權的歷史檔案 |
| 安全說明 | 有外部服務、憑證或私人資料的程式，要說清楚資料去向、保存方式與回報管道；可連到既有 SECURITY／privacy 文件 | 靜態首頁不需要虛構 runtime 安全章節；不得宣稱完美去敏 |
| 授權與追蹤檔 | 檢查 LICENSE、依賴／素材授權、實際 tracked／staged／打包內容；缺授權選擇不可擅自選 MIT。ignore 規則涵蓋此專案產物與私人 runtime | .gitignore 不會移除已 tracked 檔案，也不能取代掃描 |
| 結構與 metadata | 保留有用途的入口／設定；檢查 manifest、skill frontmatter、references 和實際被使用的 scripts。若 lock 檔影響語言統計，可用 .gitattributes 標記 generated |
根目錄 AGENTS、CONTRIBUTING、SECURITY、CHANGELOG 都可能合理；純 Mermaid 不需建立空 images 目錄 |
| 安裝與測試 | 對程式發布驗證實際安裝／建置入口和相關測試；用隔離環境。現有必要 checks 應通過；既有失敗與本次回歸分清楚，不隱藏或略過 required checks | 文件／profile 更新以連結與渲染為主，不硬加無關程式測試；離線成功不代表 live 整合成功 |
| CI | 核對現有 workflow 是否真的測試預定 artifact／平台；需要持續驗證的程式發布可補 CI | 靜態文件不一定需要新 workflow；存在 YAML 不代表跑過 |
| Docker | 專案提供或宣稱支援 Docker 時，才驗 build、compose 與必要 health check；測試不得占用正式服務或憑證 | 原生平台依賴不是 Docker-ready；無容器交付就不補 Dockerfile／badge |
| 命名與 commit | 沿用 repo／user 的命名和 commit 慣例；檢查 metadata 是否含私人資訊 | 普通作者署名、Co-Authored-By 不等於洩漏；不得為統一樣式刪署名或重寫歷史 |
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.
- 1mo ago Changed 7e25dc9f39da
- 1mo ago Changed · -173 lines · -35 tokens per session 6df7d7f3527b
- 1mo ago First seen · 280 lines · 99 tokens per session scan A ce96bce8908d
prep-repo is a skill published in the GitHub repository KerberosClaw/kc_ai_skills (79 stars, last pushed 13d ago), licensed MIT. It adds 64 tokens to every session and 3,506 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-08-30.
Other skills, from other repositories
pypi-release
This skill should be used when releasing tunacode-cli to PyPI. It keeps the existing local release checks, then hands the actual PyPI upload to a GitHub Actions workflow that uses the repository's PYPIAPITOKEN secret.
github-repo-care
Protocol for safely creating, publishing, releasing, auditing, and maintaining GitHub repositories: check local rules and locks, create .gitignore before the first add, run privacy checks, prepare README/i18n/banner/metadata, verify release tags and GitHub releases, and update organization profiles, llms.txt files…
migrate-rename
Evolutionary file renaming with wrapper files, plus the terse MOVED-stub standard for whole folder/file moves. Enables renames without hard breaks — references are organically updated through usage.
reissverschluss-merge
Reißverschluss-Verfahren (englisch: zipper merge) für sehr konfliktreiche Merges: Wenn zwei divergente Branches oder ein PR und sein Zielbranch BEIDE wertvolle, unvereinbar erscheinende Änderungen tragen, wird nicht pauschal eine Seite gewählt, sondern Abschnitt für Abschnitt verglichen, per Entscheidungstabelle…
cut-release
Decide whether unreleased develop work warrants a release, pick the version per semver.org, and cut it following this repo's established commit/PR/tag/release pattern.
github
GitHub via gh CLI: PRs, issues, reviews, repos, auth.