atlas-databricks
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and directs the user to download configuration and dependency files from official vendor domains such as https://atlasgo.io.\n- [REMOTE_CODE_EXECUTION]: The documentation includes an installation command that fetches a shell script from the vendor's official source at https://atlasgo.sh and executes it locally to install the required tool flavor.\n- [INDIRECT_PROMPT_INJECTION]: The skill describes processes where the agent ingests database metadata and migration file content, which introduces a surface for indirect prompt injection if those sources are compromised.\n
- Ingestion points: Database schema metadata from information_schema and content from local migration files processed by the CLI tool.\n
- Boundary markers: The instructions do not define specific prompt delimiters or boundary markers for separating external data from agent instructions.\n
- Capability inventory: The Atlas CLI tool provides capabilities for executing SQL statements, writing migration files, and performing network operations against the Databricks API.\n
- Sanitization: The skill does not provide specific guidance on sanitizing or validating ingested metadata before it is processed by the agent.
Audit Metadata