reflect-paper
Pass
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains no malicious code, obfuscation, or exfiltration patterns. It follows best practices by including guardrails for handling confidential content and avoiding the creation of fabricated information.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external research drafts and metadata. While this presents an inherent attack surface for indirect prompt injection, the skill includes instructions to keep sensitive content local and verify external sources, which reduces the practical risk.
- Ingestion points: SKILL.md (Inputs section) mentions LaTeX source, PDF, markdown, Word text, and bibliography files.
- Boundary markers: None explicitly defined in the provided markdown instructions.
- Capability inventory: The skill references tools for network reading (
fetch-paper), literature searching (find-papers), and report generation. - Sanitization: No explicit sanitization or filtering logic is provided in the instruction set.
Audit Metadata