attio-deploy-integration
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill enforces secure secret management by instructing the agent to inject tokens at runtime rather than hardcoding them in manifests or images.\n- [SAFE]: All external references are restricted to official Attio documentation, which is a well-known service.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted deployment manifests and logs while possessing capabilities to write to the file system.\n
- Ingestion points: Deployment manifests, secret names, and health checks are read via Read, Glob, and Grep in SKILL.md.\n
- Boundary markers: Includes explicit 'Approval Boundaries' requiring human verification for production writes and schema changes.\n
- Capability inventory: Utilizes Write and Edit for artifact updates and WebFetch for network access.\n
- Sanitization: No programmatic sanitization is defined, instead relying on human approval for the rollout process.
Audit Metadata