frontend-review-testing
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local shell script located at
scripts/audit-coverage.sh. This script is part of the vendor's package and is used to analyze repository coverage data as part of the intended audit functionality. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted project data including
coverage-summary.json,package.json, and various configuration files to generate a testing report. This creates a potential vector for indirect prompt injection if those files contain malicious instructions. - Ingestion points: The agent reads
coverage/coverage-summary.json,package.json,vitest.config.*, andplaywright.config.*from the target repository. - Boundary markers: The instructions do not define explicit delimiters or instructions to ignore embedded commands within the analyzed files.
- Capability inventory: The skill has the ability to execute a local shell script and write report files to the filesystem.
- Sanitization: No specific sanitization or validation of the external file content is requested before the agent processes it.
Audit Metadata