gitlawb

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The skill involves managing issues and pull requests on a decentralized git network, which allows outsiders to submit untrusted text via issues, pull requests, and tasks.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill documentation instructs the agent to install tools and plugins via external script and registry URLs (https://gitlawb.com/install.sh, @gitlawb/gl, @gitlawb/opencode) belonging to an unofficial, arbitrary project domain without strong first-party provenance or integrity checks, constituting a high-risk operational dependency.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 0.70). The skill documentation describes operations on a decentralized git network including creating and approving bounties with escrow release, which constitutes a direct financial operation (bounty creation and approval with financial settlement/escrow release). The access control relies on cryptographic DIDs and node/UCAN registration tokens (gl register, gl identity new), but lacks independent third-party or execution-layer authorization checks beyond the agent's own signed credentials/private keys. Therefore, it falls under direct operations without qualifying independent access control (High severity).