company-brain-company-qa
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from multiple external sources, including SQL databases, ticket exports (JSON/CSV), and document folders. This creates a surface for indirect prompt injection where malicious content within those files could influence the agent's behavior.
- Ingestion points: SQL database tables, ticket export files, and local document folders.
- Boundary markers: None specified in the instructions.
- Capability inventory: Executes local Python scripts via
uv run, reads local files, and connects to external databases. - Sanitization: No explicit sanitization or filtering of the ingested content is mentioned beyond the internal processing by the
cogneelibrary. - [COMMAND_EXECUTION]: The skill facilitates the execution of local Python scripts using the
uvtool. - Evidence: Instructions include commands such as
uv run python examples/cookbooks/company_brain/company_qa/company_qa.pyanduv run cognee-cli forget. - Context: These scripts are part of the repository's cookbooks and are intended for processing the user's provided data.
Audit Metadata