skills/openclaw/clawhub/convex-ship/Gen Agent Trust Hub

convex-ship

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill initiates the convex deploy command to push the project backend to the cloud.
  • [EXTERNAL_DOWNLOADS]: The skill interacts with the Convex platform's official moderated publishing infrastructure and the *.convex.app domain.
  • [CREDENTIALS_UNSAFE]: The instructions specify that the agent does not hold the deployment key, using a secure sign-in process for cloud access instead.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user project files (Ingestion point: project source in SKILL.md; Capability: command execution and publishing in SKILL.md; Boundary/Sanitization: none), but includes a mandatory user confirmation step to mitigate risks from potentially malicious content in the processed data.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 06:16 PM