owasp-zap
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches official OWASP ZAP Docker images from ghcr.io/zaproxy/zaproxy and GitHub Actions from the zaproxy organization. These are well-known and official sources for the Zed Attack Proxy project.\n- [COMMAND_EXECUTION]: Runs security scanning scripts (zap-baseline.py, zap-full-scan.py) via Docker to perform web security audits. This is the primary and intended purpose of the skill.\n- [INDIRECT_PROMPT_INJECTION]:\n
- Ingestion points: Ingests external target URLs and API specification files for scanning.\n
- Boundary markers: Includes warnings that report artifacts can contain sensitive data like cookies or tokens and should be treated as sensitive.\n
- Capability inventory: Docker command execution and local file writing for reports (SKILL.md).\n
- Sanitization: Relies on the security tool's internal handling of web content; documentation advises manual triage of results to confirm findings.\n- [PRIVILEGE_ESCALATION]: Uses chmod a+w on a local directory to ensure the Docker container has the necessary permissions to write report artifacts, which is standard practice for mounting volumes in this context.
Audit Metadata