gap-aware-search-protocol
Pass
Audited by Gen Agent Trust Hub on Jun 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is entirely instructional and defines a multi-step process for identifying and targeting knowledge gaps during research. No malicious code, network exfiltration, or safety bypasses were found.
- [INDIRECT_PROMPT_INJECTION]: The protocol involves retrieving data from external sources (Step 4). While this represents an attack surface for indirect prompt injection from retrieved academic papers, the risk is inherent to any search capability and the skill does not include any instructions that would weaken the agent's safety filters when processing that data.
- Ingestion points: External search results retrieved from ArXiv and other sources in Step 4.
- Boundary markers: Not present in the instruction set.
- Capability inventory: No executable capabilities (subprocess, eval, etc.) are present in the skill file.
- Sanitization: No specific sanitization or validation logic is defined for the retrieved search results.
Audit Metadata