anti-defensive-writing-en
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists entirely of markdown-based instructions for stylistic and narrative adjustments to academic papers. It does not include scripts, shell commands, or configuration for external tool access.
- [SAFE]: There are no detected patterns for data exfiltration, credential theft, or unauthorized file system access. The skill operates purely within the text-processing context.
- [SAFE]: No obfuscation, remote code execution, or persistence mechanisms are present.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of user-provided academic manuscripts.
- Ingestion points: User-supplied text triggered by prompts such as "polish my paper" or "revise manuscript".
- Boundary markers: Not present; the skill does not define specific delimiters for user input.
- Capability inventory: The skill lacks any dangerous capabilities; it has no access to the network, file system (beyond reading the input), or shell execution.
- Sanitization: No sanitization is performed on the input text.
Audit Metadata