skills/bluebagai/skills/mcp-builder/Gen Agent Trust Hub

mcp-builder

Pass

Audited by Gen Agent Trust Hub on Feb 21, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The file scripts/connections.py enables the execution of local commands using the mcp library's stdio_client. This is the standard mechanism for local MCP server communication and is required for the skill to function as intended for testing servers.
  • [EXTERNAL_DOWNLOADS] (SAFE): The skill instructs the agent to fetch documentation from the official Model Context Protocol website and its associated GitHub repositories. These are reputable sources for protocol specifications and SDK details.
  • [DATA_EXFILTRATION] (SAFE): Analysis of the provided Python and Markdown files shows no hardcoded credentials, access to sensitive file paths (like ~/.ssh), or unauthorized data transmission to external endpoints.
  • [PROMPT_INJECTION] (SAFE): Indirect Prompt Injection Surface Analysis: 1. Ingestion points: SKILL.md fetches remote documentation from modelcontextprotocol.io and GitHub. 2. Boundary markers: Absent. 3. Capability inventory: scripts/connections.py provides local command execution (stdio) and network capabilities (HTTP/SSE). 4. Sanitization: Absent. Although this creates an ingestion surface for untrusted data, the specific sources are official protocol resources, and the risk is considered safe within the scope of a development guide.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 21, 2026, 08:22 AM