agentleFS
alexnodelandalexnodeland/hoist/.claude/skills/check/SKILL.md

check

Skill2 starsChanged 33 days agoMITView on GitHub

Run exactly what Hoist's CI runs (formatting, clippy, tests and doctests, rustdoc, and the end-to-end walkthrough job when it applies) and interpret any failure. Use before committing or pushing, before opening or updating a pull request, or to reproduce a failed CI job locally.

88 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use daily-arxiv, exp-eval and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • LambenthanSkill

    daily-arxiv

    每日从 arXiv 拉取新论文,过滤相关性,auto-ingest 高优先级论文,检测 SOTA 更新

    Also in 5 projects that use this

  • LambenthanSkill

    exp-eval

    实验判决门:Review LLM 独立评判实验结果 → 4 种判决路径 → 自动更新 claims confidence、ideas status、graph edges

    Also in 5 projects that use this

  • skyllwtSkill

    exp-pilot-eval

    Pilot result evaluation — read pilot results, apply success criteria, update idea page (pilot_result, failure_reason if failed), generate PILOT_VERDICT_REPORT. Called by /ideate Phase 5 after /exp-pilot-run.

    Also in 4 projects that use this

  • skyllwtSkill

    exp-pilot-run

    Pilot experiment execution — read Pilot Spec YAML, write pilot code, run experiment(Confirm with the user before operation and require the applicant to conduct manual inspection), return results. Called by /ideate Phase 5. Does NOT modify wiki pages or judge pass/fail.

    Also in 4 projects that use this

  • JuliusBrusseeSkill

    backprop

    Bug → spec protocol. When a bug is found or a test fails, trace the cause, decide whether a new §V invariant would catch recurrence, append to §B. This is the one non-obvious thing SDD does that plan-then-execute doesn't. Triggers on test failure, bug report, post-mortem, or explicit user ask.

    Also in 4 projects that use this

  • +6 more from Lambenthan
  • +1 more from skyllwt
  • Other versions88 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 88

    1. tw93/claude-health★ 7.6ksame name
    2. tw93/Waza★ 7.6ksame name
    3. tokio-rs/topcoat★ 7.2ksame name
    4. katanemo/archgw★ 7.1ksame name
    5. katanemo/plano★ 7.1ksame name
    6. Giskard-AI/giskard-oss★ 6ksame name
    7. VibiumDev/vibium★ 3ksame name
    8. skyllwt/OmegaWiki★ 1.7ksame name
    9. DAIR-AutoResearch/AutoSci★ 1.7ksame name
    10. PKU-DAIR/AutoSci★ 1.7ksame name

    78 more not listed

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository5 other files it gives its agents.

The file

Files
SKILL.md

Check before pushing

CI is .github/workflows/ci.yml (jobs check, ui, msrv, walkthrough and platforms) and .github/workflows/coverage.yml, on every pull request and push to main. container.yml runs when the image's inputs change; pages.yml builds the site only after a merge. Run what applies to the change, in this order, and fix the first failure before going on.

1. Always: the check job

make ci    # make check lint test types doc

That is cargo fmt --all -- --check, cargo clippy --all-targets -- -D warnings (and again on hoist-console --features ts), cargo test --all-targets, cargo test --doc, the TypeScript drift test (cargo test -p hoist-console --features ts --lib api::typescript) and RUSTDOCFLAGS="-D warnings" cargo doc --no-deps, as CI runs them.

The drift test rewrites console/src/api/generated/ when a type the API serves changed: commit what it wrote, and it passes on the next run.

2. When the proxy, the CLI, an example or a script changed: the walkthrough job

cargo build -p hoist-report -p hoist-mcp
BIN=${CARGO_TARGET_DIR:-target}/debug
sh examples/quickstart/run.sh --bin "$BIN"         # no key, no network
sh packaging/test-install.sh                        # install.sh against a local release
shellcheck install.sh examples/quickstart/run.sh packaging/test-install.sh packaging/homebrew/fill.sh
python3 scripts/walkthrough.py --bin "$BIN"         # needs Node: npx fetches the MCP filesystem server
python3 scripts/http_check.py --bin "$BIN"          # needs Node: the reference Streamable HTTP server
python3 -m doctest scripts/telecom_workflow.py
python3 -m doctest scripts/ceiling.py
bash -n scripts/bench/replay.sh
python3 scripts/bench/tables.py --work scripts/bench/fixture --json /tmp/rows.json

make quickstart and make walkthrough run the first and the fifth. CI also asserts one cell of the fixture's table; the command is in ci.yml. Skip shellcheck if it is not installed, and say so.

3. When console/ or the console's API changed: the ui job

make ui-check    # formatting, lint, types, unit tests and the build
make e2e         # Playwright in Chromium: the mock API, with axe-core's accessibility rules on every page, then the console on tests/fixtures/home

A change to a type the API serves needs make types first, and the TypeScript it writes committed. In a cloud session the session hook installs console/node_modules; Playwright's Chromium is at PLAYWRIGHT_BROWSERS_PATH when the machine provides one, and npx --prefix console playwright install chromium fetches it otherwise.

4. When it applies

  • Dependencies, Cargo.lock or a newer std API: make msrv checks the workspace on Rust 1.88 with the lockfile (it installs that toolchain, about 550 MB).
  • Any change to crates/: make coverage fails under COVERAGE_MIN in the Makefile, and when a line the branch adds since its merge base with origin/main is not run by any test (git fetch origin main first). The instrumented build needs about 4 GB of disk.
  • website/, docs/, paper/, site/, brand/ or CHANGELOG.md: make site, which fails on a dead link. CI does not build the site before the merge.
  • Platform-specific code: the platforms job builds on macOS and Windows, which cannot be run here. Keep Unix-only APIs behind #[cfg(unix)].
  • The Dockerfile or what goes into the image: make docker, then the checks in container.yml's build job.

Reading failures

FailureWhat it meansWhat to do
cargo fmt prints a diffCode is not formattedmake fmt
A clippy lintWarnings are errorsFix the code. Add an #[allow] only with a comment saying why
docs/cli.md is out of date for hoist (or hoist-mcp, hoist-procedure)A command, option or help text changedmake bless, read the diff of docs/cli.md, commit it
docs/review.md is out of dateflow-show or flow-diff output changedmake bless, read the diff, commit it
docs/formats.md does not name [...]A flow or arbiter file gained a fieldDocument the field in docs/formats.md
A rustdoc error such as links to private item or an unresolved linkThe public docs link to what they cannot reachFix the path, or make a private item a code span
MSRV: rustc 1.88 is not supported by the following packagesA locked dependency needs a newer RustPin that dependency to a release that supports 1.88 (cargo update -p NAME --precise VERSION), or raise rust-version in Cargo.toml and the msrv job's toolchain together, with a CHANGELOG line
Coverage under the thresholdTests cover less than COVERAGE_MINAdd tests. Lower the threshold only with a reason in the commit
path:line: added, and no test runs itA line the branch adds is not run by any testAdd a test that runs it, or remove the line. A branch that cannot happen is better removed than tested
quickstart: FAILED: ...The loop no longer does what examples/quickstart/README.md saysThe message names the step. Fix the code, or the README and the script together
scripts/walkthrough.py failsThe code and docs/walkthrough.md disagreeChange them together: CI exists so they cannot drift
npx cannot fetch a serverNo network hereReport it as unverified; CI runs it

Report what ran, what passed, and anything not run, with the reason.

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.