regression-cicd-hooks

Fail

Audited by Gen Agent Trust Hub on Jun 18, 2026

Risk Level: HIGHCREDENTIALS_UNSAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [CREDENTIALS_UNSAFE]: The Gitea Actions example instructions include commands to read sensitive local credential files (~/.config/gitea/token) to authenticate API requests.
  • [EXTERNAL_DOWNLOADS]: The skill's Gitea integration template downloads baseline artifacts from an unverified external domain (git.integrolabs.net).
  • [COMMAND_EXECUTION]: The skill automates the creation and modification of a local Git pre-commit hook (.git/hooks/pre-commit) to execute shell-based regression checks.
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection (Category 8):
  • Ingestion points: Reads test and performance results from JSON files (test-results.json, performance-results.json) generated during the build process.
  • Boundary markers: No explicit delimiters or instructions are provided to the agent or CI environment to ignore malicious content within the test data when generating reports.
  • Capability inventory: The skill uses GitHub, GitLab, and Gitea APIs to post comments containing data derived from the ingested test results to Pull Requests and Merge Requests.
  • Sanitization: There is no evidence of sanitization or escaping of the ingested test data before it is interpolated into the PR/MR comment templates.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
Jun 18, 2026, 05:35 PM
Security Audit — agent-trust-hub — regression-cicd-hooks