shared-task-workflow

Pass

Audited by Gen Agent Trust Hub on Apr 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely instructional and defines a procedural workflow. It does not contain any scripts, binary executables, or shell commands.
  • [DATA_EXFILTRATION]: No network-enabled tools or operations are requested or used. The workflow remains within the local environment provided to the agent.
  • [PROMPT_INJECTION]: The skill explicitly instructs the agent to activate 'shared-ai-safeguards' and mandates user confirmation before any code changes, reducing the risk of autonomous malicious actions.
  • [CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or sensitive environment variable patterns were found in the instructions.
  • [COMMAND_EXECUTION]: There are no shell commands or subprocess calls defined in this workflow orchestrator.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 3, 2026, 01:57 PM