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 yimwoo/codex-agenteam --skill ci-repairgit clone --depth 1 https://github.com/yimwoo/codex-agenteamWrote 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/yimwoo/codex-agenteam/ci-repair)<a href="https://agentmods.dev/skills/yimwoo/codex-agenteam/ci-repair"><img src="https://agentmods.dev/badge/skills/yimwoo/codex-agenteam/ci-repair/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/yimwoo/codex-agenteam/ci-repair"><img src="https://agentmods.dev/badge/skills/yimwoo/codex-agenteam/ci-repair.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 | $0.00026 | $0.02036 |
| Opus 5 | $0.00013 | $0.01018 |
| Sonnet 5 | $0.00005 | $0.00407 |
| Haiku 4.5 | $0.00003 | $0.00204 |
Grade A, and why
ci-repair 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 10d 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 — 253 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AgenTeam CI Repair
Fix a CI failure on a pull request. Fetches the failure context from GitHub Actions, dispatches the dev role with bounded logs, verifies the fix locally, and pushes only if verification passes.
GitHub Actions only. The workflow is a bounded, resumable monitor → diagnose → repair → verify loop.
Loop Contract
- Defaults:
max_repair_cycles: 3,max_elapsed_minutes: 90, and a five-minute CI polling interval. User-supplied lower budgets win. - Persist every cycle in
.agenteam/ci-repair/pr-<pr_number>.json, keyed byhead_sha. Record the run ID, failed jobs, repair attempt, local verification, pushed SHA, last heartbeat, and stop reason. - Emit a heartbeat while waiting for a role or CI. If the host cannot remain active, schedule a follow-up/automation and stop cleanly instead of holding a silent blocking wait.
- Before replaying a repair after restart, compare the checkpoint head SHA and current PR head. Never repeat a completed repair for the same SHA.
- Stop explicitly on success, exhausted cycle/elapsed budget, changed PR head, unavailable evidence, or required human approval.
Process
1. Accept Input
Get the PR reference from the user:
- PR number:
$ateam:ci-repair #42 - Branch name:
$ateam:ci-repair feature/add-auth - Nothing: use the current branch
2. Resolve PR
Resolve the input to a canonical tuple: pr_number, pr_url,
pr_branch, head_sha.
If PR number given:
gh pr view <number> --json number,url,headRefName,headRefOid
If branch given or current branch:
BRANCH=$(git rev-parse --abbrev-ref HEAD) # if no input
gh pr list --head <branch> --json number,url,headRefName,headRefOid --limit 1
If no PR found: "No PR found for branch <branch>. Push and create
a PR first, then retry." Stop.
Extract:
pr_numberfromnumberpr_urlfromurlpr_branchfromheadRefNamehead_shafromheadRefOid
3. Check CI Status
Get the latest CI run for the PR's current head commit:
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.
- 10d ago First seen · 253 lines · 26 tokens per session scan A 6705126f61f4
ci-repair is a skill published in the GitHub repository yimwoo/codex-agenteam (13 stars, last pushed 2mo ago), licensed MIT. It adds 26 tokens to every session and 2,036 once invoked, about $0.0001 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 2026-08-30.
Other skills, from other repositories
lookup-warn
Get context warnings for a file based on its extension and name. Surfaces known gotchas before the user hits them. Triggers on: warn me, gotchas, pitfalls, what should I watch out for, check this file.
rpi-fix-ci
Diagnose existing CI failures and repair their root causes locally with regression evidence and complete verification before any authorized publication.
overllm
Find the LLM call you didn't need — your GPT call is a regex. Static, zero-config linter that flags unnecessary GPT/Claude/AI API calls a regex, the stdlib, or a library already does (parsing dates, extracting emails/JSON, sorting, classifying with fixed labels). Runs locally, no model, no API key, nothing phones…
langbot-plugin-dev
Develop, debug, and test LangBot plugins. Use when creating new LangBot plugins, fixing plugin bugs, setting up a LangBot test environment, or testing plugins via WebSocket. Covers plugin component architecture (EventListener, Command, Tool), the plugin SDK API (invokellm, getllmmodels, sendmessage, plugin storage)…
baby-sit
Monitor a GitHub pull request until CI is green, diagnose failures, and rerun only evidence-backed flaky GitHub Actions jobs.
ax-cpp-agent-observability
Use when writing C++ code with axllm for agent tracing, centralized and multi-tenant usage accounting, action logs, runtime diagnostics, replay, and production debugging.