tw-research-citation-checker
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes arbitrary citation lists provided by users and incorporates this data into search queries for academic databases and web searches. This creates an attack surface where instructions hidden within a reference could attempt to manipulate the agent's logic.
- Ingestion points: The user-provided citation list requested in Step 1 of SKILL.md.
- Boundary markers: The skill logic does not specify the use of delimiters or 'ignore' instructions to isolate untrusted user data.
- Capability inventory: The skill employs WebSearch and the Consensus MCP to interact with external data sources.
- Sanitization: No validation or sanitization procedures are documented for the incoming citation text.
Audit Metadata