mindmap
Pass
Audited by Gen Agent Trust Hub on May 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions direct the agent to execute shell commands using the
$ARGUMENTSvariable (e.g.,uv run python3 $CLAUDE_PROJECT_DIR/tools/brain.py brief $ARGUMENTS). This pattern is vulnerable to command injection if the agent does not properly sanitize the user-provided input before shell execution. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting data from external sources.
- Ingestion points: Data is read from the
recon/directory and the output of thebrain.pyandintel_engine.pytools. - Boundary markers: The instructions do not specify any delimiters or safety markers to differentiate between data and instructions when processing these inputs.
- Capability inventory: The skill has the capability to execute shell commands across all included scripts via
uv run python3. - Sanitization: There is no evidence of sanitization or validation of the ingested data before it is incorporated into the agent's context.
Audit Metadata