c-level-advisor
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a robust and secure architecture for multi-agent strategic advisory. It follows defensive programming principles, particularly in its handling of data and inter-agent communication.
- [SAFE]: Data protection is a core feature. The anonymization protocol (found in context-engine/references/anonymization-protocol.md) explicitly requires stripping sensitive financial figures, customer names, and employee identities before any data leaves the local environment. All company context is stored locally in
~/.claude/company-context.md. - [SAFE]: The 25 included Python tools are verified to be dependency-free, using only the Python standard library. They perform legitimate analytical tasks consistent with the skill's purpose, such as financial modeling and operational bottleneck analysis.
- [SAFE]: The inter-agent protocol (
[INVOKE:role|question]) includes strict loop prevention and depth limits, ensuring the agent system remains stable and predictable. - [SAFE]: Indirect prompt injection risks are addressed through structured data ingestion (JSON/CSV) and clear boundary markers in the agent interaction logs.
Audit Metadata