regression-selection-agent
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data including
impact_map.md,affected_components.json, and various test files found in the repository. - Ingestion points: The agent reads files from
.agile-v/phases/01-impact/and performs a broad search for test files throughout the repository (e.g.,.spec.ts,_test.py). - Boundary markers: The instructions lack explicit boundary markers or specific prompts to ignore embedded instructions within the ingested data files.
- Capability inventory: The skill is authorized to search the filesystem, read arbitrary files, and write output files to the
.agile-v/phases/03-regression/directory. - Sanitization: There are no documented sanitization, escaping, or validation steps for the external content before it is processed by the agent.
Audit Metadata