map-observable-boundaries
Pass
Audited by Gen Agent Trust Hub on Sep 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a provided local Python script (
scripts/check-boundaries.py) to perform structural validation of the generated mapping files. This is a standard functional use of local command execution.\n- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to analyze external artifacts such as API definitions, logs, and system schemas to catalog system boundaries. While these constitute external data ingestion, the instructions focus on structural mapping rather than autonomous control flow, posing a low risk.\n - Ingestion points: External artifacts including event logs, API definitions, and runbooks mentioned in SKILL.md.\n
- Boundary markers: Absent. The skill does not define specific delimiters for these inputs.\n
- Capability inventory: Local file system access and execution of the included
check-boundaries.pyscript.\n - Sanitization: The
check-boundaries.pyscript performs structural JSON validation to ensure required fields are present and correctly formatted.
Audit Metadata