with one click
lint
Run all local linters — markdown, code smells, and Mermaid diagrams
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Menu
Run all local linters — markdown, code smells, and Mermaid diagrams
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Write Bash commands that don't trigger unnecessary permission prompts. Use Read/Edit/Grep instead of head/tail/sed/awk/cat/echo; split chained commands (&&, ;, ||) into separate Bash calls; drop diagnostic suffixes like `; echo "EXIT=$?"`. The allowlist matches whole command strings — every mismatch costs the user attention.
Commit user-named files atomically via scripts/commit-pathspec.sh (which wraps `git commit -m "..." -- <files>` and writes the provenance token the pre-commit hook validates). Applies the CLAUDE.md "Pre-commit hook failures on unrelated changes" protocol when the hook fails. Never adds --no-verify silently — explicit user approval is required.
Scaffold a new epic file in docs/developers/tasks/open/ and run housekeep
Scaffold a new idea file in docs/developers/ideas/open/ and regenerate the ideas OVERVIEW.md
Set a task to active — updates status, moves to active/, runs housekeep. Also resumes paused tasks (paused/ → active/ or active/ with closed prerequisites).
Mark a task as closed — writes effort_actual, moves the file to closed/, runs housekeep, and commits
| name | lint |
| description | Run all local linters — markdown, code smells, and Mermaid diagrams |
Run all local linters and report findings. This mirrors the checks performed by CI.
Steps:
Markdown — run make lint-markdown (auto-fixes issues in place via markdownlint-cli2).
Report which files were changed, if any.
Code smells — run python scripts/check_code_smells.py.
Report any functions exceeding 60 lines, files exceeding 400 lines,
nesting depth beyond 4 levels, or leftover TODO/FIXME/HACK/XXX comments.
Mermaid diagrams — run python scripts/validate_mermaid.py.
Report any diagrams that fail validation.
After all three checks, print a summary:
Markdown: OK (or: N files fixed)
Code smells: OK (or: list of issues)
Mermaid: OK (or: list of failures)
If markdown auto-fixed files, remind the user to review and stage the changes. Do not run clang-format or clang-tidy — those require a connected device build environment.