regression-proof
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external bug reports, source code, and existing test suites as defined in Step 1. This creates a vulnerability surface where malicious code or instructions embedded in the ingested data could influence the agent's behavior.
- Ingestion points: The agent establishes context from the request, relevant code, and existing tests in Step 1.
- Boundary markers: Absent. The skill does not specify delimiters or instructions to ignore embedded commands within the processed data.
- Capability inventory: The agent is instructed to perform command line execution to run baselines, regression tests, and adjacent checks in Steps 1, 2, 4, and 5.
- Sanitization: Absent. No mention of validating or sanitizing external content before executing tests.
- [COMMAND_EXECUTION]: The skill requires the agent to execute shell commands to run baselines and tests (Steps 1, 2, 4). While necessary for the skill's purpose, this capability is the primary vector for potential exploitation via the indirect prompt injection surface identified above.
Audit Metadata