x402

[Skill Scanner] Credential file access detected No direct indicators of malware or obfuscated/hidden backdoors are present in this skill manifest. The functionality is coherent with the stated purpose: discovering paid endpoints and making USDC payments. However, the feature set is high-impact: it probes arbitrary URLs with multiple HTTP methods and can sign/send USDC payments once the user is authenticated. The main supply-chain and security risks are operational: (1) reliance on a remote CDP/bazaar index (which could be compromised and direct users to malicious endpoints), (2) potentially unsafe probing of user-specified URLs (side-effects), and (3) lack of documented guarantees about how wallet credentials are stored or whether signing is performed locally. I assess this as not overtly malicious but as carrying meaningful security risk that requires careful implementation and user safeguards (explicit confirmations, local-only signing, index integrity verification). LLM verification: The SKILL.md describes a legitimate-seeming CLI tool for discovering and paying for paid APIs using the x402 protocol. There is no direct evidence of malicious code or obfuscation in this documentation fragment. However, several high-risk behaviors are documented: automatic monetary payments, automatic probing of arbitrary URLs, and local caching without described integrity checks. The absence of concrete endpoints or implementation code prevents verification that payments and auth are routed on