testing-reality-checker
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute several shell commands (
ls,grep,cat) and a local shell script (./qa-playwright-capture.sh). These commands are used to verify project file structures, search for specific UI implementation strings (e.g., 'glass', 'morphism'), and trigger a Playwright-based screenshot capture process. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data generated during the testing process, specifically from
test-results.json. This data is used to inform the agent's 'Realistic Quality Certification' and populate the final report, creating a potential vector where malicious content in the tested application could influence the agent's findings. - Ingestion points: The agent is instructed to read content from
public/qa-screenshots/test-results.jsonas specified in Step 1 and Step 2 of the mandatory process. - Boundary markers: The instructions do not define delimiters or specific 'ignore' directives for the data read from external test files.
- Capability inventory: The skill possesses shell execution capabilities and uses the gathered data to make binary decisions (PASS/FAIL) and generate reports.
- Sanitization: There is no evidence of sanitization or validation of the test result content before it is processed by the agent's reasoning engine.
Audit Metadata