company-brain-follow-up-agent
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill reads and processes data from external untrusted sources, specifically Gmail and Linear. This introduces a risk where malicious content embedded in an email or issue could influence the agent's behavior during the follow-up process.
- Ingestion points: Ingests data from Gmail and Linear into the 'cognee' dataset (referenced in SKILL.md).
- Capability inventory: Executes local Python scripts via 'uv run' and posts results to Slack.
- Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore potentially malicious instructions within the ingested text.
- Sanitization: There is no mention of sanitizing or escaping the data pulled from external sources before it is processed by the LLM.
- [COMMAND_EXECUTION]: The skill instructs the agent to run local Python scripts using the 'uv' package manager. This includes passing dynamic arguments such as '--days' and '--clear' to the execution environment.
Audit Metadata