Back to skill
Skillv2.1.1
VirusTotal security
Autonomous Agent Skills · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
SuspiciousApr 29, 2026, 3:20 AM
- Hash
- 8323ea2c1886339604b5a363bd4c5eaa4e99b07234c278378ba389abb7759b0d
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: autonomous-agent Version: 2.1.1 The skill bundle is classified as suspicious due to its inherent high-risk capabilities, including autonomous transaction signing and interaction with multiple external blockchain-related services. The `src/agent/tools/localTools.js` and `src/lib/wallet.js` files handle the generation and local storage of private keys with `chmod 0o600`, which are then used by `src/lib/aptos/signPayment.js` and `src/lib/evm/signPayment.js` to sign transactions for payment to `https://x402-navy.vercel.app/facilitator` (configured in `src/run-agent.js`). While these actions are central to the skill's stated purpose of enabling x402-paid MCP tools via `https://arnstein.ch`, the significant financial autonomy granted to the agent, coupled with the potential for prompt injection to manipulate tool arguments or payment objectives, presents a notable risk. Additionally, the nested Moltbook skill (`.moltbot/skills/moltbook/HEARTBEAT.md`) includes a self-update mechanism via `curl` from `https://www.moltbook.com`, which could pose a supply chain risk if the update source were compromised.
- External report
- View on VirusTotal