datahub-setup
Pass
Audited by Gen Agent Trust Hub on May 27, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the Bash tool to perform environment checks, manage Python virtual environments, and execute DataHub CLI commands. These operations are restricted to the primary purpose of the skill and are performed using standard, safe practices such as local virtual environments.
- [EXTERNAL_DOWNLOADS]: The skill installs the official 'acryl-datahub' package via pip. This is the legitimate CLI package for the DataHub project, and the skill correctly avoids global installation or the use of elevated privileges (sudo).
- [DATA_EXPOSURE]: The skill interacts with the local configuration file '~/.datahubenv' to manage connection settings. It includes explicit instructions to redact sensitive tokens from the user's view and prevent them from being logged, mitigating the risk of credential exposure.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface where it accepts user-provided configuration values (URLs, tokens, and profile names).
- Ingestion points: User input gathered in Phase 1 (Step 3: Configuration) and Phase 2 (Step 5: Preferences).
- Boundary markers: The skill uses markdown placeholders in templates ('templates/agent-config.template.md') and specific security rules in 'SKILL.md' to separate configuration data from instructions.
- Capability inventory: The skill uses 'Bash' for package installation (pip) and connectivity testing (datahub CLI), and performs file system writes to save configuration profiles.
- Sanitization: The 'SKILL.md' includes mandatory rules to validate GMS URLs and mask tokens, ensuring that untrusted input does not compromise the agent's output or security posture.
Audit Metadata