Chealth — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Chealth (Plugin) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
Aggregate score unchanged between these scans.
The primary manifest — the file an agent reads to learn what this artifact does.
CLI tool and Claude Code plugin to check CLAUDE.md health against Anthropic best practices.
From within a Claude Code session, add the marketplace and install:
/plugin marketplace add danielithomas/chealth
/plugin install chealth@danielithomas-chealthOr use the interactive plugin browser:
/pluginNavigate to the Discover tab after adding the marketplace.
Then run /chealth inside any Claude Code session. The plugin runs deterministic CLI checks followed by semantic analysis powered by Claude.
Requires Python 3.10+ and uv.
git clone https://github.com/danielithomas/chealth.git
cd chealth
uv sync# Check current directory
uv run chealth .
# Check a specific path
uv run chealth /path/to/project
# JSON output (for automation)
uv run chealth --format json .
# Run a specific check
uv run chealth --check line-count .Platform wrappers are provided in bin/:
# Bash
./bin/chealth.sh .
# PowerShell
.\bin\chealth.ps1 .| # | Rule | Severity |
|---|---|---|
| 1 | Keep it under 200 lines | WARN/DANGER |
| 2 | Include only universally applicable content | Semantic |
| 3 | Use markdown structure | WARN |
| 4 | Modularise with imports | Semantic |
| 5 | Document build and test commands | WARN |
| 6 | State architectural decisions and key file locations | Semantic |
| 7 | Document repo etiquette | Semantic |
| 8 | Use WHAT/WHY/HOW framework | Semantic |
| 9 | Be concrete, not vague | Semantic |
| 10 | Use emphasis keywords | INFO |
| 11 | Avoid contradictions | Semantic |
| 12 | Encode Explore > Plan > Code > Commit | Semantic |
| 13 | Instruct incremental commits | Semantic |
| 14 | Specify test-run preferences | Semantic |
| 15 | Use subdirectory CLAUDE.md files | Semantic |
| 16 | Use ~/.claude/CLAUDE.md for personal preferences | Semantic |
| 17 | Check project CLAUDE.md into version control | WARN |
| 18 | Treat it like a prompt — iterate | Semantic |
| 19 | Use # shortcut to build as you work | Semantic |
| 20 | Progressive disclosure | Semantic |
Checks marked Semantic are evaluated by Claude during plugin use, not by the CLI.
MIT
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.