literature-verifier
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill connects to well-known academic services and databases (e.g., CrossRef, DOI.org, CNKI, Wanfang, Baidu Scholar) to retrieve metadata for verification. These external requests are essential to the skill's primary function and target established academic infrastructure.
- [COMMAND_EXECUTION]: The skill uses local Python scripts to perform network queries and parse metadata from academic search results. The scripts are focused on specific verification tasks and do not execute arbitrary shell commands.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: The skill processes user-supplied citation text and external web content retrieved from academic search result pages via MCP browser tools or URL verification scripts.
- Boundary markers: Instructions require the agent to cross-check extracted metadata (authors, titles, journals) against the user's specific claims, providing a structured verification process that mitigates the risk of blindly following instructions embedded in search results.
- Capability inventory: The skill utilizes network requests via Python's standard libraries and browser automation tools (MCP) for navigating and reading academic sites.
- Sanitization: Search parameters are appropriately sanitized using URL encoding in the supporting Python scripts to prevent malformed or malicious queries.
Audit Metadata