find-key-internal-sources
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFE
Full Analysis
- [Data Ingestion Surface]: The skill is designed to process content from internal messaging, CRM systems, and document stores. This is a standard and necessary function for its purpose, but it inherently means the agent will encounter and interpret untrusted data from these sources.
- [Controlled Output and Actions]: The skill follows a 'read-only' principle for its initial output. Any subsequent actions, such as posting to communication channels or sharing handoff notes, require explicit user review and approval of the generated draft, preventing automated exfiltration or unauthorized messaging.
- [Identity and Freshness Validation]: To ensure reliability, the skill implements logic to exclude inactive accounts and prioritizes signals from the last 90 days when identifying experts, which reduces the risk of referencing stale or compromised internal metadata.
- [Cross-Skill Integration]: The skill utilizes a preflight mechanism by referencing a separate configuration skill (
sales:user-context). This modular approach ensures consistent application of organizational policies and source-mapping rules.
Audit Metadata