Skip to content
Back to skills

Lint

ASecurity

Use when finding and fixing the project's configured lint findings in the current session, without a separate MEOW fixer agent.

  • 2 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added October 26, 2026
developmentbashrefactoring

Works with

  • cli

Security analysis

A100/100

Scanned October 26, 2026

npx -y skills add maya8237/MEOW --skill lint --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Lint?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Lint
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/maya8237-lint/badge)](https://www.skillsdirectory.com/skills/maya8237-lint)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: lint
description: Use when finding and fixing the project's configured lint findings in the current session, without a separate MEOW fixer agent.
---

# lint

Runs the project's `[[lint]]` commands and has **you** — this session — fix what
they report, using `meow native lint` for the facts. Read
[`../_shared/native-mode.md`](../_shared/native-mode.md) for helper conventions. Never run plain
`meow run --lint-fix` here: that spins up meow's own fixer agent, duplicating
your work and racing your edits. Dependency or install problems: [dependencies](../_shared/dependencies.md).

## Native mode

1. The project needs `.meow/config.toml` with at least one `[[lint]]` entry;
otherwise tell the user and stop (point at `docs/INTEGRATIONS.md` for the
`[[lint]]` schema, or invoke `/meow:onboard` to configure it).
2. Run `meow native prepare --no-worktree --allow-dirty --work-dir "<project-path>"`
   for `max_rounds` and `docs_dir`, then
   `meow native round <docs_dir>/lint-fix.md --reset --work-dir "<project-path>"`.
3. Auto-fix pass, then check: `meow native lint --fix --all-blocking --work-dir "<project-path>"`.
   `--all-blocking` fixes every configured command unconditionally, matching
   `meow run --lint-fix`'s own CLI behavior -- `gate` only controls what fails a
   *review*, not what this skill leaves alone.
   - `"clean": true` -> report that nothing is left and stop.
   - Otherwise call `meow native round <docs_dir>/lint-fix.md --work-dir
     "<project-path>"` before the first manual fix attempt; stop if exhausted.
     Each entry in `blocking` is a `$ <command>` block of raw linter
     output. Before editing, run `meow native prompt lint-fixer --active-dir
     <active_dir> --work-dir "<project-path>"` and follow its returned `system_prompt`, just as the CLI
     lint-fixer role follows `lint_fixer_prompt`. Fix each finding yourself
     with the smallest edit that resolves it, without unrelated refactoring;
     run `meow native lint --file <path> --active-dir <active_dir>` after each
     edit. Do not launch another agent.
4. Re-check with `meow native lint --all-blocking --work-dir "<project-path>"`; call
   `meow native round <docs_dir>/lint-fix.md --work-dir "<project-path>"` before each further fix attempt
   and stop if it says `exhausted`.
5. Report the final outcome: clean, or what is still outstanding and why (e.g. a
   project-wide type error needing a design decision).

## CLI mode

Report-only, then fix by hand:

```bash
meow run --lint-fix --report-only --work-dir "<project-path>"
```

Exit 0 with `Lint is clean -- no issues found.` means done; otherwise fix each
`$ <command>` block it printed and re-run to confirm.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.