work-on-linear-issue
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTION
Full Analysis
- CREDENTIALS_UNSAFE (LOW): The skill requires a
LINEAR_API_KEYand suggests it can be stored in a.envfile. While no secrets are hardcoded, the use of environment variables or.envfiles for sensitive keys is a common target for data exposure findings. - COMMAND_EXECUTION (LOW): The skill executes a local script (
work-on-linear-issue) that utilizes system utilities likecurlandjq. This is an expected execution surface for the skill's functionality. - Indirect Prompt Injection (LOW): (Category 8)
- Ingestion points: Fetches external issue details, descriptions, and comments from Linear.
- Boundary markers: None specified in the instructions to help the agent distinguish between Linear data and system instructions.
- Capability inventory: Access to
curl,jq, and broad instructions to 'Explore the codebase'. - Sanitization: No mention of sanitizing or escaping the data fetched from the Linear API before the agent reviews it.
- Note: An attacker could place malicious instructions inside a Linear issue to influence the agent during the planning phase.
Audit Metadata