specialist-specialized-mcp-builder

Pass

Audited by Gen Agent Trust Hub on Aug 15, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill defines a specialized developer persona for building and testing MCP servers. The instructions focus on production-quality software development practices, including Zod validation, error handling, and structured outputs, with no evidence of malicious intent or unauthorized access patterns.
  • [PROMPT_INJECTION]: The skill uses string interpolation to inject user input into a subagent prompt. Although this is a standard design for template-based agents, the absence of explicit boundary markers or sanitization logic technically exposes an indirect prompt injection surface.
  • Ingestion points: The {the user's question} placeholder in SKILL.md.
  • Boundary markers: Absent; user input is not isolated from system instructions using delimiters or XML tags.
  • Capability inventory: Spawns ephemeral specialist agents using the Agent tool.
  • Sanitization: No evidence of input filtering or escaping was found in the provided file.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 15, 2026, 08:50 PM
Security Audit — agent-trust-hub — specialist-specialized-mcp-builder