neural-mechanism-research
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: No security vulnerabilities or malicious patterns were detected in the analyzed skill files. The content is strictly instructional and focused on scientific research methodology.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface by design, as it processes user-provided queries and hypotheses about neural networks.
- Ingestion points: User research questions and proposed technical explanations are ingested and processed according to instructions in SKILL.md.
- Boundary markers: The instructions recommend quoting user wording and reconstructing the query, which helps delineate untrusted input from the agent's logic.
- Capability inventory: The skill is purely instructional and does not define or request access to sensitive capabilities such as network exfiltration, file system modification, or shell execution.
- Sanitization: No specific input sanitization is implemented, as the skill focus is on analytical reasoning and reporting.
Audit Metadata