research-paper-review
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its primary function of processing untrusted external data.
- Ingestion points: Processes user-provided PDF, LaTeX source files, and plain text papers (SKILL.md, Step 1). It also attempts to fetch guidelines from external 'venue websites' (SKILL.md, Step 0).
- Boundary markers: None identified. There are no explicit instructions to the agent to ignore instructions or malicious payloads embedded within the research papers or fetched website content.
- Capability inventory: Uses file reading tools ('Read' tool with pages parameter, 'Grep', 'Glob') and potentially network operations if 'try to find them on the venue website' involves automated browsing or fetching (SKILL.md, Step 0 & 1).
- Sanitization: No evidence of sanitization or validation of the content extracted from papers or external websites before it is used to influence the agent's review logic.
Audit Metadata