Linear Agents (AIG, Signals, Interaction)
Pass
Audited by Gen Agent Trust Hub on Aug 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains instructional content and code examples for legitimate integration with Linear's API. All external links and API endpoints reference 'linear.app', which is a well-known service. The TypeScript snippets demonstrate standard OAuth and GraphQL usage patterns.
- [CREDENTIALS_SAFE]: The documentation mentions authentication requirements (e.g., 'agent_oauth_token', 'Linear-Actor-Token') but correctly instructs users to capture or mint these tokens through secure backends rather than hardcoding them in the skill.
- [PROMPT_INJECTION]: The skill describes an architecture where an agent receives 'interaction events' such as user mentions and issue comments from the Linear platform. This represents a potential surface for indirect prompt injection, as the agent may process untrusted data from external users. However, the skill focuses on API signaling and does not include unsafe instruction interpolation.
- Ingestion points: Linear webhook events for agent interactions (mentions, assignments).
- Boundary markers: Not explicitly defined in the API-focused code snippets.
- Capability inventory: The skill is configured with 'Read', 'Grep', 'Glob', and 'Bash' tools, and demonstrates GraphQL API writing capabilities.
- Sanitization: Not provided in the documentation snippets.
Audit Metadata