scholar-anti-ai-writing
Pass
Audited by Gen Agent Trust Hub on Jul 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a well-structured tool for academic prose improvement. It operates locally on user-provided files and does not exhibit any signs of malicious intent, credential theft, or unauthorized network activity.
- [COMMAND_EXECUTION]: The skill utilizes common CLI tools like Bash, Grep, and Glob. These are leveraged for administrative tasks such as file discovery and verifying the structural integrity of LaTeX documents during the rewriting process. No arbitrary or unsafe command construction was identified.
- [PROMPT_INJECTION]: The skill processes user-supplied text, which constitutes a potential indirect prompt injection surface (Category 8).
- Ingestion points: User-specified file paths or pasted text ingested in Phase 1 (phases/01-detect-score.md).
- Boundary markers: No explicit delimiters or boundary markers are defined for the input text.
- Capability inventory: The skill has permission to read, write, and edit files, and execute shell commands (Bash).
- Sanitization: There is no mention of sanitizing or escaping the input text before the agent processes or rewrites it.
- Context: This surface is inherent to the skill's function as a text editor and represents a low risk in its intended academic use case.
Audit Metadata