Skip to content
Get packs

Skill Trust Checker

Scores another agent skill's design, safety and fit before you install or deploy it

Source Download skill
Repository
anthropics/claude-for-legal
Stars
9,870
Licence
Apache-2.0
Files
1 file

What it does

This skill makes Claude review another skill before you trust it. It checks the skill against a 13-point legal design framework, looks for prompt-injection and hidden-instruction patterns, and gives a verdict in one of three bands: Ready, Some Concern, or Material Concerns. It also checks whether the skill overlaps with skills you already have installed. The file shown is only the SKILL.md, so the full dependency check cannot be completed from it.

Great for

  • Legal or operations teams deciding whether to install a community skill
  • Admins reviewing a first-party skill before rolling it out to a team
  • Anyone who asks 'should I trust this skill?' before installing

Skip it if

  • Guaranteeing a skill is safe: the scan is an AI heuristic, not a security audit, and a clean result is not an approval
  • Reviewing non-legal skills against legal-specific failure modes, which may produce less relevant findings

Try asking

  • “Should I trust this community skill before I install it? [paste SKILL.md]”
  • “Is this skill well-designed? Check the folder at ./skills/contract-review”
  • “Review this updated version of the skill for any new security-sensitive changes”
Step-by-step guide: how to use Skill Trust CheckerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the Apache-2.0 licence and an attribution note.

    skills-qa.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip skills-qa.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip skills-qa.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Provide the full skill folder, not just SKILL.md, so hooks, commands and agents can be checked for automatic triggers
  • Treat a clean verdict as one input; still read any flagged text and any diffs yourself before approving
  • Re-run the check on every new version, since the skill can change after you first trust it

What's inside

  • SKILL.md

Redistributed unmodified from anthropics/claude-for-legal under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.