Skills
skills/zerodeploy-dev/skills/rollback/Gen Agent Trust Hub

rollback

Pass

Audited by Gen Agent Trust Hub on Feb 25, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the zerodeploy command-line interface to manage deployments. The use of this vendor-specific tool is expected behavior for the skill's stated purpose.
  • [INDIRECT_PROMPT_INJECTION]: The skill retrieves organization and site details from a local zerodeploy.json file. * Ingestion points: Reads from zerodeploy.json. * Boundary markers: Not present, but disable-model-invocation: true restricts model behavior. * Capability inventory: Execution is limited to zerodeploy commands. * Sanitization: Relies on the external CLI for input validation.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 25, 2026, 06:00 PM