paper-context-resolver
Pass
Audited by Gen Agent Trust Hub on May 29, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: No malicious patterns, unauthorized access attempts, or suspicious instructions were detected. The skill is purely informational.
- [NO_CODE]: The skill does not include any executable scripts, binaries, or source code. It is composed of markdown guidelines and YAML configuration.
- [EXTERNAL_DOWNLOADS]: The skill directs the agent to locate information from academic sources such as arXiv, OpenReview, and Google Scholar. These are well-known, trusted platforms for research data, and the skill does not automate the download or execution of files from these sources.
- [INDIRECT_PROMPT_INJECTION]: The skill creates a surface for ingesting untrusted data by instructing the agent to read external research papers (ingestion points:
references/paper-assisted-reproduction.md). There are no specific boundary markers or sanitization logic defined in the instructions; however, the risk is negligible as the skill lacks code execution capabilities.
Audit Metadata