databricks-config
Pass
Audited by Gen Agent Trust Hub on Mar 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
databricksCLI to perform authentication, interpolating user-provided values like the profile name directly into shell commands. This pattern can lead to arbitrary command execution if the input is not strictly validated. - [CREDENTIALS_UNSAFE]: The skill reads and modifies
~/.databrickscfg, which contains sensitive authentication tokens. Although the skill includes instructions to redact these tokens in the agent's response, the skill itself possesses full access to the credentials. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. Ingestion points: User-provided parameters
profile_nameandworkspace_hostinSKILL.md. Boundary markers: None identified. Capability inventory: Subprocess execution viadatabricks auth loginand filesystem access to~/.databrickscfg. Sanitization: There is logic to extract strings from URLs, but no explicit sanitization for profile names used in command arguments.
Audit Metadata