atlas-review
Pass
Audited by Gen Agent Trust Hub on Nov 3, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process an externally provided Atlas node to drive file system operations and command execution.
- Ingestion points: The skill ingests prose sentences, file paths, and line numbers from a Project Atlas node.
- Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the node content as untrusted or to ignore instructions embedded within the claims.
- Capability inventory: The agent is instructed to use git rev-parse, open specific files at designated lines, and run grep searches based on the input data.
- Sanitization: The instructions do not specify any validation or sanitization for the paths or search terms extracted from the input node.
Audit Metadata