run-parallel-agents-feature-build

Pass

Audited by Gen Agent Trust Hub on Mar 16, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes user-provided task lists and interpolates them into subagent instructions, creating a surface for indirect prompt injection.
  • Ingestion points: User requests and task lists analyzed in Step 1 of the Core Workflow.
  • Boundary markers: Absent; the brief template does not employ delimiters or ignore-instructions warnings for user data.
  • Capability inventory: Uses the Agent tool for task execution, which possesses file modification and command execution capabilities.
  • Sanitization: Absent; the workflow does not include validation or escaping of external task descriptions.- [SAFE]: The skill utilizes environment-provided isolation features (worktree) to maintain state integrity during parallel execution.- [SAFE]: No credentials exposure, malicious network behavior, or obfuscation techniques were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 16, 2026, 01:03 AM