agentleFS
Zwin-uxZwin-ux/kit/.claude/skills/kit-ci-triage/SKILL.md

kit-ci-triage

Skill0 starsChanged 16 days agoMITView on GitHub

Get build and test proof for Kit and diagnose CI failures, especially when working on Windows. Use before claiming a Kit change works, when a check fails on a PR, or when a local cargo build behaves oddly on Windows.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Encoding: read and write text as UTF-8 explicitly (encoding="utf-8" in Python).

    line 34
  2. Distinctive:

    Add or fix the test that would have caught it.

    line 40

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Kit CI triage

Where proof comes from

CheckWorkflow / jobRuns onProves
fmt, clippy -D warnings, cargo test --workspacerust.yml → qualityubuntu, macos, windowsevery crate, the TUI snapshots, the CLI integration tests
Packagingrust.yml → crates.io packageubuntucargo package --workspace --locked; version copies agree
Cold startrust.yml → startup budgetubuntu, windowskit --version mean under 100 ms (hyperfine)
npm launcherrust.yml → npm package (scripts/npm-smoke.mjs)3 OSesthe packed tarballs install and kit runs through the shim
Installersinstallers.yml3 OSesinstall.sh / install.ps1 against a local release

rust.yml is path-filtered: a docs-only PR runs no Rust checks. That is not a pass for code; it means there was no code. Read results with gh pr checks <n> -R Zwin-ux/kit and gh run view <run-id> -R Zwin-ux/kit --log-failed. Use the app's PR status tools rather than polling. Say which rows ran and on which OS before calling anything verified.

Known traps

  • Windows path length: a checkout in a deep folder makes link.exe fail with LNK1104: cannot open file … build_script_build-….exe. Set CARGO_TARGET_DIR to a short folder (Git Bash: export CARGO_TARGET_DIR=C:/kit-target; PowerShell: $env:CARGO_TARGET_DIR = 'C:\kit-target').
  • NoDefaultCurrentDirectoryInExePath=1 (set on the owner's hardened machine): cmd /C name.cmd no longer finds name.cmd in the current folder. Tests and checks must write .\name.cmd.
  • CRLF: core.autocrlf=true checks files out with \r\n. Strip \r before grepping, and assert the count of any scripted replacement. Insta snapshots compare with line endings normalised.
  • Insta: a changed screen writes .snap.new and fails. Run INSTA_UPDATE=always cargo test -p kitctl-tui (PowerShell: $env:INSTA_UPDATE = 'always'; cargo test -p kitctl-tui) only after reading the diff, then commit the .snap and read git diff of it line by line.
  • Clippy on newer stable: CI uses dtolnay/rust-toolchain@stable, which can be newer than your local toolchain. A lint that only CI reports is real; fix it, don't pin the toolchain down.
  • Encoding: read and write text as UTF-8 explicitly (encoding="utf-8" in Python). ©, ·, → and box-drawing characters appear in Kit's output and docs.

Failure loop

  1. Find the failing job and the first real error line, not the summary.
  2. Reproduce locally if the toolchain allows; otherwise reason from the log and push a fix to the PR branch.
  3. Add or fix the test that would have caught it.
  4. Push and read the new run on every OS before saying it is fixed.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.