The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the datahub CLI tool for metadata and lineage operations. It includes explicit security instructions to validate user inputs (URNs and search terms) and reject shell metacharacters, effectively mitigating command injection risks.
[PROMPT_INJECTION]: The skill handles data from the DataHub ecosystem. It maintains security by scoping operations to the legitimate datahub tool and providing clear instructions to fallback if advanced features like sub-agent dispatch are unavailable. It does not attempt to override agent safety protocols.
[SAFE]: The skill's behavior is consistent with its stated purpose of data lineage analysis. No evidence of obfuscation, hardcoded credentials, or unauthorized network operations was found.

datahub-lineage