manual-tester
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it ingests untrusted test designs, plans, and case IDs provided by users or projects. \n
- Ingestion points: Test design and plan documents in the project repository (SKILL.md, references/test-plan-intake.md).\n
- Boundary markers: The skill relies on a procedural intake and classification process to audit conditions rather than technical delimiters.\n
- Capability inventory: Access to Write, Edit, and Bash tools (curl, psql, docker, jq, python3) which could be targeted by malicious instructions.\n
- Sanitization: The instructions focus on human review and classification of external content to mitigate risks, without explicit string sanitization.\n- [COMMAND_EXECUTION]: The scripts/test_bootstrap_adapter.py and scripts/test_resolve_customization.py files use subprocess.run to execute the skill's own utility scripts during development testing, which is a benign and standard development activity.\n- [EXTERNAL_DOWNLOADS]: Configuration scripts reference the tomli package as a standard fallback dependency for TOML parsing in Python environments older than version 3.11.
Audit Metadata