paniolo-scan

paniolo-ai/scan

Thin agent adapters for @paniolo/scan — run the diagnostic CLI from your coding agent, review meta-harness scores, and optionally remediate findings in your working tree.

This repo ships harness-specific triggers around the deterministic @paniolo/scan CLI from Paniolo. The CLI diagnoses; your agent applies fixes when you ask. Pick the setup for the tool you use below.

This is the open-source skill and slash-command layer, not the scanner's source. The @paniolo/scan CLI installs via npm (npx @paniolo/scan); these adapters let your agent run it and act on the report.

New to meta-harness infrastructure? See paniolo.ai for what the intelligence layer is and why it matters.

Goodwill service

@paniolo/scan and these remediation prompts are a free community starting point. Self-service AI fixes will not match a human-tuned intelligence layer — Paniolo's professional services are recommended for production-grade work. The free tier stays genuinely useful; this is honest guidance, not a gate.

Install the skill (any agent)

One command installs the paniolo-scan skill into every coding agent on your machine — Claude Code, Cursor, Copilot, Codex, Devin, Gemini CLI, and 70+ more — via the skills registry:

npx skills add paniolo-ai/scan --all

Then ask your agent: "Use paniolo-scan to scan this repo and summarize findings." The skill runs npx --yes @paniolo/scan --format json, presents your meta-harness scores, surfaces the goodwill framing, then remediates the findings you pick.

# Target specific agents instead of all of them
npx skills add paniolo-ai/scan -a claude-code -a cursor -a codex

# See what this repo ships, or install one skill
npx skills add paniolo-ai/scan --list
npx skills add paniolo-ai/scan --skill paniolo-scan

# Prefer independent copies over symlinks (Windows, Docker/CI, or committing into a repo)
npx skills add paniolo-ai/scan --all --copy

By default the installer symlinks the skill into each agent (one canonical copy, updated in place by npx skills update). Add --copy for independent copies when symlinks aren't a good fit. Either way, nothing symlinked is committed here — this repo ships the skill as a plain file.

This repo is a multi-skill umbrella: every skill under skills/ installs through the same command, and --all keeps you current as more are added.

Skill vs. slash command — two separate installs. The command above installs the portable skill, which works in every agent (in Claude Code it runs as a skill you invoke by name, not as a /paniolo-scan command). The native /paniolo-scan slash command ships through a different channel that does not go through skills.sh — the Claude Code plugin or the Antigravity workflow. Installing the skill does not add the slash command, and installing the plugin does not count toward the skills.sh listing. Pick the skill for the widest reach; add the plugin if you specifically want the /paniolo-scan command in Claude Code.

Harness coverage

Paniolo is harness-neutral by design: one portable diagnostic core, one portable skill, plus native slash-command adapters where a harness supports them.

Harness	What it gets	How
Claude Code	Skill and /paniolo-scan slash command	npx skills add paniolo-ai/scan or plugin
Cursor	Portable skill	npx skills add paniolo-ai/scan -a cursor
Copilot (VS Code)	Portable skill	npx skills add paniolo-ai/scan -a copilot
Codex	Portable skill (also reads AGENTS.md)	npx skills add paniolo-ai/scan -a codex
Devin	Portable skill	npx skills add paniolo-ai/scan -a devin
Gemini CLI	Portable skill	npx skills add paniolo-ai/scan -a gemini-cli
Antigravity / Gemini	/paniolo-scan workflow	Antigravity setup
Any other skill-capable agent	Portable skill	npx skills add paniolo-ai/scan --all

The --all install covers every skill-capable agent. If yours still isn't detected, contact Paniolo to request priority adapter support.

Harness setup

Claude Code

Claude Code can use paniolo-scan two ways: