zentao-tour

Pass

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing zentao-cli via common package managers (npm, pnpm, bun). This tool is the official CLI for the ZenTao platform and is maintained by the skill's author, easysoft.
  • [COMMAND_EXECUTION]: The agent is instructed to use zentao-cli to perform CRUD (Create, Read, Update, Delete) operations on ZenTao project data. These actions are performed to simulate real-world workflows for the user during the tour.
  • [CREDENTIALS_UNSAFE]: The skill mentions the storage location of user credentials in the ~/.config/zentao/zentao.json file. This is standard configuration information for the zentao-cli tool and is used to verify connectivity and authentication during the setup process.
  • [PROMPT_INJECTION]: The skill processes data retrieved from the ZenTao server, creating a potential surface for indirect prompt injection.
  • Ingestion points: Data enters the agent context through zentao task, zentao bug, and zentao story command outputs (referenced in roles/dev.md, roles/test.md, and roles/pm.md).
  • Boundary markers: The instructions do not define specific delimiters or warnings to ignore instructions embedded in the project data.
  • Capability inventory: The skill possesses write capabilities through zentao create, update, resolve, and close commands across all role-based scripts.
  • Sanitization: There is no explicit sanitization or filtering of the content retrieved from the external ZenTao instance.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 12, 2026, 03:57 AM
Security Audit — agent-trust-hub — zentao-tour