skills/wenjunduan/rlues/agent-teams/Gen Agent Trust Hub

agent-teams

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill configuration manages automated environment setup via worktree-init.cjs and executes testing suites using specialized sub-agents. These actions are performed within isolated worktrees as part of the intended development workflow.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it relies on external data in plan.md to coordinate sub-agent activities.\n
  • Ingestion points: plan.md (orchestration source).\n
  • Boundary markers: No explicit delimiter or instruction-ignore markers are defined in the configuration.\n
  • Capability inventory: Sub-agents like builder, validator, and e2e-runner have filesystem access and command execution permissions within their worktree environments.\n
  • Sanitization: The architecture implements verification through validator and security-auditor agents, which acts as a downstream check but does not sanitize the initial input.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 03:28 PM