SCAN
Your skill ZIP is scanned against OWASP LLM Top 10. Score below 70 = refused.
score ≥ 70 required🦞 OPENCLAW INTEGRATION
Protect your OpenClaw skills revenue from copies
You built a skill that people install and run.
AgentVerif makes sure no one can copy or tamper with it after the first download.
Once your skill is published, you lose control
Anyone can copy your skill, modify it, and republish it.
Without a certificate, your users can’t tell the original from a fake.
Tampered skills damage your reputation — and you’re the one they blame.
SCAN → SIGN → VERIFY
SIGN
A cryptographic hash is generated. SIGNATURE.json is injected. A License ID is issued.
AC-XXXX-XXXXVERIFY
Users verify at verify.agentverif.com or via CLI. Tampered skills are flagged instantly.
VERIFIED / TAMPEREDWorks inside your OpenClaw workflow
Via CLI
pip install agentverif-sign
agentverif-sign sign ./my-skill.zip
# ✅ SIGNED — License: AC-XXXX-XXXX
Via OpenClaw skill
npx clawhub@latest install agentcop
/security sign ./my-skill.zip
/security verify AC-XXXX-XXXX
# ✅ Protected by AgentVerif
What OpenClaw creators get
Your users trust your skill
The ✅ badge answers “is this safe?” before they ask. Verified skills get installed more.
Tamper detection built in
Modified copies are caught instantly. If the hash doesn’t match, the skill won’t run.
Revoke instantly
Kill a malicious copy with one command. The revoked version stops verifying everywhere.
Works everywhere
CLI, GitHub Action, OpenClaw skill, MCP. One certificate — valid across every runtime.
Your brand on every command
✅ Protected by AgentVerif
Every output from your skill automatically shows this line.
Your users see it on every scan and verify.
Your reputation grows with every command.
First 100 OpenClaw skills creators get free access
No credit card. Just feedback.
Zero risk — if it doesn’t protect your skill, you lose nothing.
Signed yours? Open an issue titled “First 100: [Your License ID]” — we feature you on the leaderboard.