Your agent can sign itself up.
Security for AI agents. Your agent registers, scans itself, and gets a signed score in about 60 seconds. You approve it with an email code.
Paste this into your agent:
Read https://securemyagent.ai/skill.md and follow itWorks with OpenClaw, Claude Code, Hermes, Codex, Cursor, n8n and any agent that can read a URL and run a command.
The supply chain your agent installs from is poisoned.
Personal agents hold real keys and run real code. Nobody sells protection to the person running one agent on a Mac mini. Figures from vendors are approximate.
- 341 of 2,857
- audited ClawHub skills were malicious. 335 installed the AMOS stealer. ThaiCERT
- 13.4%
- of 3,984 agent skills had a critical security issue. Snyk ToxicSkills
- 5
- malicious skills passed both VirusTotal and ClawScan. Unit 42
- 21.6k to 135k
- OpenClaw gateways exposed on port 18789. CVE-2026-25253 is a one-click RCE, CVSS 8.8. clawdocs
- 1.5M
- agent tokens exposed by one Supabase key in client JavaScript. Infosecurity Magazine
- 30+
- MCP CVEs in Jan and Feb 2026. About 43% were shell injection. Cycode
Four modules. The first is free.
| Module | Free | Pro |
|---|---|---|
| CheckupA local scan of your agent. Signed 0-100 score with a fix for each finding. | On demand. Open source. Runs on your machine. | Continuous monitoring through the heartbeat. |
| Verdict RegistryIs this skill or MCP server safe at this exact version? Rug-pull diffing on every release. | 100 lookups a day. Public. | Unlimited lookups. Alerts when an installed version turns bad. |
| GuardA local policy enforcer between your agent and its tools. Egress allowlist, secret-exfiltration block, risky-tool approval. | Not included. | Claude Code hook, OpenClaw plugin, MCP proxy. |
| AttestationA signed statement that your agent is secure. Other agents verify it with one call. | Not included. | 30-day attestation, badge, auto-renewed by heartbeat. |
Advisories, a threat feed filtered to what your agent has installed, come with every plan.
Pricing
Launch pricing
Free
$0
1 agent. Checkup on demand. 100 registry lookups a day. Public advisories.
Pro
$9
per agent per month
Monitoring, Guard, Attestation and badge, alerts, unlimited lookups, sma fix --vault.
Team
$49
per month for 10 agents
Everything in Pro. Shared dashboard, ClawGRC export, policy packs. $4 per extra agent.
One of four products an agent can own.
Identity, secrets, security and governance. One DID links an agent across all four.
- IdentityOpenEntity + Moltbot DenA portable DID and signed attestations.
- SecretsAgent Secret StoreCredentials your agent uses without seeing.
- Securitysecuremyagent (you are here)Proof it is safe, kept current, verifiable by anyone.
- GovernanceClawGRCAudit trail and compliance mapping for owners and teams.