mcp-ops

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill consists entirely of educational documentation and code snippets. No active malicious logic or security vulnerabilities were identified.
  • [PROMPT_INJECTION]: The skill describes patterns for handling untrusted input through MCP tool arguments.
  • Ingestion points: Code handlers in SKILL.md and references/tool-patterns.md receive input via the arguments dictionary.
  • Boundary markers: Patterns emphasize the use of inputSchema to define expected data structures.
  • Capability inventory: Demonstrated patterns include network requests via httpx (auth-patterns.md) and local file system persistence via aiosqlite (state-patterns.md).
  • Sanitization: The skill explicitly promotes the use of Pydantic for strict input validation and sanitization as seen in references/tool-patterns.md.
  • [DATA_EXFILTRATION]: The skill includes a pattern for local data persistence using a SQLite database located at ~/.my-mcp-server/state.db. This is a standard practice for maintaining application state and does not involve unauthorized data transmission.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 10:05 PM