agent-ci-dev

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: Executes CI lifecycle commands using pnpm and specialized background shell tools provided by the platform.
  • Evidence: Uses pnpm agent-ci-dev run --all -q -p to start the CI suite and pnpm agent-ci-dev retry for targeted re-runs.
  • [PROMPT_INJECTION]: Processes CI output logs to detect failure patterns and coordinate retries, which creates a surface for indirect prompt injection from the code being tested.
  • Ingestion points: Reads up to 200 lines of the CI output file (SKILL.md).
  • Boundary markers: None provided for log content.
  • Capability inventory: Can execute shell commands via bash_background (SKILL.md).
  • Sanitization: No sanitization of log content is specified.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 02:24 PM