flow-test
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes instructions from an external file to determine browser actions, creating a potential vector for manipulation if the file source is untrusted.\n
- Ingestion points: Reads flow definitions and critical flows from the local file
tests/e2e-test-plan.md.\n - Boundary markers: No delimiters or safety warnings are specified to distinguish between test data and instructions within the plan file.\n
- Capability inventory: The skill uses Playwright MCP to perform high-privilege browser actions including navigation, form submission, and execution of interaction logic.\n
- Sanitization: There is no evidence of sanitization, validation, or filtering of the content read from the test plan before it influences agent behavior.\n- [COMMAND_EXECUTION]: The skill provides a framework for the agent to execute a series of browser automation commands (navigate, click, fill_form) which, while intended for testing, grant significant control over web sessions based on instructional input.
Audit Metadata