agentic-os
Pass
Audited by Gen Agent Trust Hub on Jun 20, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The architecture utilizes a file-based state (data/) that agents read from, which creates a vulnerability surface for indirect prompt injection.
- Ingestion points: Project context and logs in the data/ folder.
- Boundary markers: None specified.
- Capability inventory: Agent roles include filesystem access, Git operations, and test runner execution.
- Sanitization: No sanitization is mentioned for data read from the filesystem.
- [COMMAND_EXECUTION]: Provides templates for persistent system services (LaunchAgent and systemd) to automate tasks. These are intentional functional components of the described framework.
- [CREDENTIALS_UNSAFE]: Appropriately identifies hardcoded credentials as an anti-pattern and recommends the use of environment variables.
Audit Metadata