daily-research-review
Pass
Audited by Gen Agent Trust Hub on Oct 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external research-method articles and web sources provided by the user, which serves as an entry point for indirect prompt injection attacks.
- Ingestion points:
SKILL.mddescribes a workflow where the user brings a "research-method article or web source into the review context" for archival or summary extraction. - Boundary markers: The instructions do not define specific delimiters or "ignore instructions" wrappers to prevent the agent from obeying malicious commands embedded in the external sources.
- Capability inventory: The skill has the capability to read and write to durable research documentation in Feishu, Notion, and Obsidian, which could be leveraged if an injection succeeds.
- Sanitization: No sanitization or filtering logic is specified for handling external data before it is integrated into the agent's context.
Audit Metadata