integrate-external-knowledge
Pass
Audited by Gen Agent Trust Hub on Nov 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted external data such as meeting notes, chat logs, and URLs, which could potentially contain malicious instructions aimed at influencing the agent's actions.
- Ingestion points: External data is gathered from conversation history, attachments, selections, and specified URLs as defined in the 'Intake' section of SKILL.md.
- Boundary markers: The instructions include a specific defensive directive to treat commands found within source materials as data only, preventing them from overriding the skill's logic.
- Capability inventory: The skill possesses the ability to read from network resources (URLs) and perform file system write operations to update workspace documentation.
- Sanitization: The 'Integration' and 'Quality Gate' sections require the agent to verify that no PII, internal URLs, or secrets are included in the generated output.
Audit Metadata