convex-improve-convex-plugin

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (medium risk: 0.30). The skill instructs the user to pipe a remotely hosted script into bash using a dynamic domain placeholder (<anteater>), which represents an unverified third-party code execution vector.

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

• Third-party content exposure detected (low risk: 0.10). The required workflow sends the local coding session transcript to an external endpoint and summarizes the structured findings returned by the service.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill fetches and executes an arbitrary remote script via curl from an unverified variable endpoint (<anteater>/send-transcript), constituting a weak operational dependency with high risk.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.80). The skill instructs the agent to download and execute an arbitrary remote script via a bash pipe (curl ... | bash), which poses a significant security risk to the machine state.