create-test-plan
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data, such as Pull Request titles, descriptions, comments, and the source code itself, to generate test scenarios.
- Ingestion points:
SKILL.md(Step 1 and Step 4) specifies fetching PR details and reading changed code to determine testing scope and requirements. - Boundary markers: The skill lacks explicit boundary markers or instructions to disregard potential malicious commands embedded within the external data it processes.
- Capability inventory: The skill is designed to generate plans that utilize
agent-browser,computer-use, and "direct terminal execution" (Step 2 and Step 6). - Sanitization: No explicit sanitization or filtering of external data is mentioned before it is interpolated into the generated test plan.
- [COMMAND_EXECUTION]: The skill generates instructions for test scenarios that specifically suggest the use of shell/terminal commands and automated browser tools.
- Evidence:
SKILL.md(Step 2 and Step 5) directs the agent to generate plans involving direct terminal execution and interactions with automated testing tools based on the analyzed code.
Audit Metadata