skills/zhangyu1818/skills/deep-search/Gen Agent Trust Hub

deep-search

Pass

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No prompt injection or safety bypass instructions were found; the skill enforces a strict verification workflow that prioritizes facts over hypotheses.
  • [SAFE]: No data exposure or exfiltration risks; the skill's operations are limited to information retrieval from public web and repository sources via legitimate MCP tools.
  • [SAFE]: No obfuscation, hidden payloads, or suspicious code patterns were detected.
  • [SAFE]: The risk of indirect prompt injection is mitigated by the design of the research protocol. Ingestion points: Brave Search, Exa, Tavily, GitHub API, and ZRead search tools. Boundary markers: Instructions explicitly mandate labeling community evidence as non-authoritative and separating navigation direction from evidence. Capability inventory: Read-only search and code analysis tools with no write or execution capabilities. Sanitization: Mandatory cross-verification against Tier A primary sources (official documentation and source code).
  • [SAFE]: All referenced external tools and services are appropriate for technical research and originate from well-known providers or the skill's author.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 12, 2026, 02:42 PM
Security Audit — agent-trust-hub — deep-search