mcp-builder
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill establishes patterns for prompt templates and documentation retrieval that ingest external data into the agent's context.
- Ingestion points: Prompt template definitions and the documentation lookup process using
mcp__context7__query-docsas described in SKILL.md. - Boundary markers: The skill recommends using markdown code blocks to delimit ingested content within prompt templates.
- Capability inventory: MCP servers designed with this skill typically involve file system access (e.g.,
readFile) and network operations via transport layers (stdio, SSE). - Sanitization: The skill emphasizes the use of Zod schemas for strict input validation, path traversal prevention, and explicitly warns against exposing stack traces to clients.
Audit Metadata