harness-code-integration
Pass
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious behavior or security violations were detected. The skill provides legitimate templates for CI/CD automation and environment configuration.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it is designed to ingest and process data from external repositories and pull requests.
- Ingestion points: Repository manifests (YAML), Helm charts, and Pull Request webhook payloads from the .harness/ directory and triggers.
- Boundary markers: No explicit delimiters or instructions were found to isolate untrusted repository content from the agent's system prompts.
- Capability inventory: The skill utilizes
Bash,Write, andWebFetchtools to execute linting, tests, and security scans within the CI environment. - Sanitization: No specific sanitization, escaping, or validation logic is defined for processing external repository content or PR metadata before interpolation into workflows.
Audit Metadata