education-paper-writer
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides structured guidance for academic writing and does not contain any executable scripts, network requests, or hardcoded credentials.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided research data and external paper search results from external skill calls to generate content. Analysis of ingestion points (user input and search results), boundary markers (absent), capabilities (text generation only), and sanitization (absent) indicates that while a surface exists, there are no exploitable capabilities such as command execution, file writing, or network exfiltration present in the skill's code.
- [METADATA_POISONING]: Metadata fields accurately reflect the skill's purpose without deceptive instructions or attempts to override agent behavior.
Audit Metadata