flisk: Skill for Claude Code

.claude/skills/flisk-ci-watch/SKILL.md

flisk-ci-watch is a skill for Claude Code from flisk-sh/flisk. It costs 85 tokens per session (795 once invoked), scanned B, original, MIT.

A workflow for watching and fixing failed checks on a GitHub pull request. Continuous integration (CI) automatically builds and tests code, and this workflow examines its failures, changes the pull request worktree, pushes fixes, and sets a notification hook.

In plain words
What is it for?
Use it to inspect failed GitHub Actions jobs, decide whether the failure belongs in the pull request, fix code or tests, rerun transient failures, and be notified when CI fails.
Why use it?
It keeps the session that opened a pull request responsible for its failing checks and helps distinguish code failures from stale or temporary service failures.

Skill for Claude Code

Written for Claude Code: installed under .claude/. Also seen: reads .claude/ paths.

This is flisk-sh/flisk's own configuration. It tells Claude Code how to work on flisk itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything flisk configures →

Reuse

Borrowing it

Nothing to install: this file belongs to flisk-sh/flisk. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/flisk-sh/flisk/main/.claude/skills/flisk-ci-watch/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/flisk-sh/flisk

Made for: Claude Code.

Wrote 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.

agentmods badge for flisk-ci-watch

README.md
[![agentmods](https://agentmods.dev/badge/skills/flisk-sh/flisk/flisk-ci-watch/github.svg)](https://agentmods.dev/skills/flisk-sh/flisk/flisk-ci-watch)
Your own site
<a href="https://agentmods.dev/skills/flisk-sh/flisk/flisk-ci-watch"><img src="https://agentmods.dev/badge/skills/flisk-sh/flisk/flisk-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.

agentmods 80×15 button for flisk-ci-watch

Your own site · 80×15
<a href="https://agentmods.dev/skills/flisk-sh/flisk/flisk-ci-watch"><img src="https://agentmods.dev/badge/skills/flisk-sh/flisk/flisk-ci-watch.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 85 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 795 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 1 finding. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00085 $0.00795
Opus 5.5 $0.00034 $0.00318
Sonnet 5.5 $0.00017 $0.00159
Haiku 4.5 $0.00009 $0.00080

Measured yesterday against content hash c22adaa27111, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade B, and why

flisk-ci-watch 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 yesterday.

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.

compatibility: Requires the gh CLI authenticated for flisk-sh/flisk and git. The optional hook needs python3 and changes ~/.claude/settings.json.
.claude/skills/flisk-ci-watch/SKILL.md · 41 lines

How it starts

The opening of the file, as written. The whole thing — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Flisk CI watch

The session that opened a pull request owns its CI. A skill only runs when it is called, so it cannot notice a failure by itself: use the steps below when you know CI is red, and the hook (last section) so the session is told.

Fix a red check

  1. See what failed, on the PR's own head: gh pr checks <number> --repo flisk-sh/flisk then gh run view --job <job-id> --log-failed. Logs are not available while the run is still in progress; wait, or read one finished job with gh api repos/flisk-sh/flisk/actions/jobs/<job-id>/logs.
  2. Compare the head the failure names with your own git rev-parse HEAD. A failure from a head you already replaced is stale: re-check, do not "fix" it.
  3. Decide whose failure it is before editing:
    • Does main fail the same check? Run the failing script on origin/main in a clean worktree. If yes, the fix belongs in its own PR against main, not in this one.
    • A Server Error from the GitHub API, or a registry 5xx, is transient: gh run rerun <run-id> --failed, no code change.
    • GEL guards / tests-with-change means a product file changed with no test file: add a test, do not exempt the file.
  4. Fix it in the worktree of that PR's branch, not in a shared checkout whose branch other sessions may switch.
  5. Run the checks for the files you touched, commit with git commit -s, and push to the PR branch. Merge main into the branch when it conflicts; never rebase or force-push.
  6. Before git commit --amend, run git status: after a merge that did nothing, an amend rewrites the previous commit and the push is rejected.

Be told when CI fails

  • Desktop app Auto-fix wakes an idle session on CI failures, merge conflicts and review comments. Turn it on per pull request. This is the one that works while you are away.
  • The failedci2claude Stop hook (https://backtogreen.flisk.si) tells the session when a turn ends and one of your open PRs has a failed check. It runs when a turn ends, so it cannot wake an idle session. It reads your PRs with one gh pr list call and reports each failure once.

Read the full file on GitHub · 41 lines

Changes

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.

  1. yesterday First seen · 41 lines · 85 tokens per session scan B c22adaa27111

Subscribe to this mod's changes

flisk-ci-watch is a skill published in the GitHub repository flisk-sh/flisk (54 stars, last pushed yesterday), licensed MIT. It adds 85 tokens to every session and 795 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-11-19.

Related

flisk-ci-watch compared with skills from other repositories

flisk-ci-watch next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
flisk-ci-watchflisk-sh/fliskSkill54yesterdayMITB85Claude Code
github-managerjpantsjoha/ai-native-developer-experienceSkill1311d agoApache-2.0A38Claude Code
docker-via-wslnetresearch/docker-development-skillSkill24yesterdaynoneA141Claude Code
deployment-patternsaffaan-m/ECCSkill307ktodayMITA41Claude Code, Kiro
agent-github-modesruvnet/rufloSkill82ktodayMITA21Claude Code
baby-sitlangchain-ai/open-sweSkill11ktodayMITA30Claude Code, Codex
chinese-git-workflowjnMetaCode/superpowers-zhSkill8.6k4d agoMITA69Claude Code