Portable agent-skill review pack
Review AI-agent skills before they cross a trust boundary.
Agent skills are becoming software packages. This kit helps teams inspect tool authority, memory writes, receipts, rollback, evals, exception behavior, and portability before installing, publishing, selling, or delegating a skill.
ZIP SHA256: 6f924696b39c840bba25bfb96ba38f09b90ace29bafe51eb09338eecfcaa8df1
Live on TechTools Launchpad as a free developer-tool listing: view listing.
30-minute verdictInstall, install with limits, ask for changes, or reject.
Permission matrixMap reads, writes, browser/API actions, spend, messages, memory, and approvals.
Receipts + rollbackRequire proof, before/after state, exception handling, and repair paths.
Directory-readyIncludes copy, captions, preview cards, and submission angles for non-X distribution.
What is inside
- 30-minute review kit
- Portable skill contract template
- Marketplace safety checklist
- Sample mini-report
- Permission manifest JSON Schema
- Example permission manifest JSON
- Browser-local permission manifest validator
- Redaction-friendly intake
- Directory/community submission pack
- Visual preview pack
- Vendor/RFP questionnaire
Preview cards
Optional custom review
For high-authority skills, the live funnel includes a redaction-friendly intake and a $29 custom mini-report.
Download the GitHub-ready repo package
OpenClaw-specific checklist
If you are reviewing an OpenClaw skill specifically, start with the OpenClaw Skill Safety Checklist before granting browser, file, memory, SaaS, message, or public-posting authority.
