testing-gh-skills
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DYNAMIC_EXECUTION]: The skill is designed to invoke local Python and Shell scripts (
testing/scripts/run-all-tests.pyandtesting/scripts/run-single-test.sh) to manage test execution and parallel worker pools. It also programmatically invokes a 'test-reviewer agent' and the 'claude' CLI tool to perform testing tasks. - [INDIRECT_PROMPT_INJECTION]: The test orchestrator reads test definitions and criteria from external markdown files and processes user requests to simulate testing scenarios. This ingestion of external data creates a potential vulnerability surface.
- Ingestion points: The orchestrator discovers and parses test scenario files located in
testing/scenarios/. - Boundary markers: There are no explicit boundary markers or 'ignore' instructions specified for the data ingested from the scenario files.
- Capability inventory: The skill has the capability to execute system commands (Python, Bash), launch parallel subprocesses, and write detailed reports to the local file system.
- Sanitization: The analysis did not find evidence of sanitization or validation routines for the content contained within the external scenario files before processing.
Audit Metadata