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 YuDefine/nuxt-supabase-starter --skill gh-ci-watchgit clone --depth 1 https://github.com/YuDefine/nuxt-supabase-starterWrote 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/yudefine/nuxt-supabase-starter/gh-ci-watch)<a href="https://agentmods.dev/skills/yudefine/nuxt-supabase-starter/gh-ci-watch"><img src="https://agentmods.dev/badge/skills/yudefine/nuxt-supabase-starter/gh-ci-watch/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/yudefine/nuxt-supabase-starter/gh-ci-watch"><img src="https://agentmods.dev/badge/skills/yudefine/nuxt-supabase-starter/gh-ci-watch.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Agent Snooping · line 59 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
- medium Agent Snooping · line 61 Skill enumerates or reads other installed skills. Access to other skills' SKILL.md files or the skills directory reveals prompt instructions, capabilities, and secrets that should be invisible to peer skills.Fix: Remove all code or instructions that list or read other skills' files or directories. Skills should operate independently; cross-skill access is a privilege escalation.
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 | $0.00077 | $0.04434 |
| Opus 5 | $0.00039 | $0.02217 |
| Sonnet 5 | $0.00015 | $0.00887 |
| Haiku 4.5 | $0.00008 | $0.00443 |
Grade A, and why
gh-ci-watch 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 today.
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 — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/gh-ci-watch — GitHub Actions 監看 / 查詢唯一入口
核心 contract：監看 CI = 機械輪詢。用 target adapter 提供的 background command runner 跑 $GH_CI_WATCH（adapter 綁定的 skill-local helper）——腳本自己 poll 到 terminal state 才 exit，主線只在完成時收到一次通知。等待期間零 LLM turn、零 token、行為 100% 確定。
Script 位置：
- Consumer 端：由 adapter 綁定
$GH_CI_WATCH - Clade home：
plugins/hub-core/scripts/gh-ci-watch.sh
機制選擇：為什麼是 background Bash + script，不是其他
這段是本 skill 存在的理由。NEVER 退回用 Agent subagent 監看 CI——那正是本 skill 要根治的事故根因。
事故實證（2026-07-25， v0.99.7 發版）：依當時規約派兩個 LLM watcher 監看 Deploy Staging / Production，實際發生四件事：(1) brief 明寫「completed 才回報」，agent 仍反覆中途回報「持續監看中…」，每次回報都是一次 LLM turn，累計 235k+ tokens 且沒有產出最終結果；(2) 監看的 staging run 被 concurrency cancel-in-progress 取消後，agent 繼續空等已死的 run；(3) 重新指派新 run id 後，watcher 口頭答應卻仍回報舊 run 結論；(4) 同一份 brief、同一個 model，兩個 watcher 行為不一致。
| 機制 | 判定 | 理由 |
|---|---|---|
| target adapter 的 background command runner + 本 script | ✅ 採用 | 官方定位就是「單次通知：告訴我 X 好了沒」。腳本達 terminal state 即 exit → 剛好一次通知；無 LLM 參與 → 零等待成本、行為確定。事故中需要「判斷力」的三件事（run 尚未建立、被 concurrency 取代、同 SHA 多條 run）其實都是機械規則，已全部編進 script（Phase 1 pending 重查、Phase 2 successor 追蹤、--since/--commit 過濾），不需要 LLM |
| LLM watcher | ❌ 禁用 | 見上方事故四點。LLM「判斷力」在這個場景是負資產：不可預測 + 每個動作燒 token。唯一例外見下方「例外」節 |
| 事件流通知器 | ❌ 不用 | CI 監看要的是恰好一次完成通知；事件流型通知若 filter 沒涵蓋所有 terminal state，crash 時沉默會跟「還在跑」一模一樣。本 script 用 RESULT: 行涵蓋全部 terminal state，從結構上排除這個坑 |
主線 wakeup / 前景 gh run watch |
❌ 不用 | 佔用主線 context / block 主線對話。gh run watch 也不處理 run 被取代 |
監看：canonical dispatch 樣板
以下命令一律由 target adapter 以 background command runner 派出（cwd = 該 repo，或帶 --repo <owner>/<repo>），派出後主線繼續原本工作，等系統的完成通知。
場景 A — 盯已知 run id
bash "$GH_CI_WATCH" run <run-id>
場景 B — 盯某 workflow 最新一條 run（push 後標準場景）
workflow 識別字串一律傳檔名（ci.yml），NEVER 傳 display name 或自己想的簡稱。
gh run list -w 只認兩種形式：workflow 檔名，或 name: 欄位的逐字 display name。
display name 是自由文字、跟檔名無關（ci.yml 的 name 常是 CI / Deploy），而且隨時可被編輯 ——
檔名要改得動 git。傳錯時 script 自 2026-08-28 起在進輪詢前就 fail fast：exit 2 並把該 repo
實際的 workflow 清單印進 RESULT: 行；先前是被當成 API 抖動重試 3 次後回通用 UNAVAILABLE，
訊息與「gh 掛了 / 沒授權」同形（ v1.272.0 實證，見
[[pitfall-gh-ci-watch-workflow-display-name-guess-fails-opaquely]]）。名字拿不準就先跑
gh workflow list，或直接用場景 A 的 run <run-id>。
What ships with it
1 file 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.
- today Changed · +3 lines 52282f419f15
- 2d ago Changed 5e419210e269
- 4d ago Changed · +3 lines aa0d6b3c7495
- 6d ago Changed dd5c5f64f6cd
- 8d ago First seen · 179 lines · 77 tokens per session scan A 9f9671cf93b5
gh-ci-watch is a skill published in the GitHub repository YuDefine/nuxt-supabase-starter (45 stars, last pushed yesterday), licensed MIT. It adds 77 tokens to every session and 4,434 once invoked, about $0.0004 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 2026-09-03.
Other skills, from other repositories
nw-par-review-criteria
Quality dimensions and review checklist for devop reviews.
pr-watch
Local PR watcher. Monitors CI status, automatically fixes failing checks by reading failure logs and applying targeted fixes, then optionally merges when all checks pass. Local CLI analog to Claude Code's cloud auto-fix feature.
start-temps-cluster
Start (or restart) a local multi-node Temps cluster using Docker-in-Docker — one control plane + 3 worker nodes, each a privileged DinD container running its own dockerd + temps agent, wired with the real multi-host overlay (VXLAN, computecidr allocation) via tools/dev-cluster/ in whichever checkout/worktree you run…
cloudflare-workers-ci-cd
Complete CI/CD guide for Cloudflare Workers using GitHub Actions and GitLab CI. Use for automated testing, deployment pipelines, preview environments, secrets management, or encountering deployment failures, workflow errors, environment configuration issues.
bitbucket
Manage repositories, pipelines, and code review with Bitbucket Cloud. Use when a user asks to set up Bitbucket repositories, configure Bitbucket Pipelines for CI/CD, manage pull requests, set up branch permissions, use Bitbucket REST API 2.0, create webhooks, manage deployment environments, set up code review…
swift-preflight
Audit a Swift / iOS / macOS repo for Xcode Cloud and TestFlight release blockers before upload - pbxproj drift, static build numbers, missing ciscripts, macOS App Store entitlements/Info.plist, headless-CI keychain tests, ad-hoc signing entitlement rejections, and flaky-UITest release gating. Reports PASS/WARN/FAIL…