cognee-install
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for setting up a virtual environment and installing the Cognee package using the
uvtool, which involves shell command execution. - [INDIRECT_PROMPT_INJECTION]: The
cognee.remember()method allows the agent to ingest data from external sources, including raw text and remote URLs. This creates a vulnerability where malicious instructions embedded in the processed data could influence the agent's behavior during subsequent operations. - Ingestion points:
cognee.remember()(accepts text, file paths, and URLs) andcognee.recall(). - Boundary markers: None identified in the skill instructions to distinguish between trusted instructions and untrusted data.
- Capability inventory: The skill facilitates reading local files and fetching content over the network for indexing into memory.
- Sanitization: The skill does not mention or implement sanitization or validation of the content retrieved from external URLs or files before processing it.
Audit Metadata