MCP_prompt_broker
by michaelprinc
Overview
Intelligent routing of prompts to optimal AI agent instruction profiles and orchestration of delegated AI tasks (Codex/Gemini CLI) in isolated Docker containers.
Installation
python -m mcp_codex_orchestratorEnvironment Variables
- OPENAI_API_KEY
- WORKSPACE_PATH
- RUNS_PATH
- SCHEMAS_PATH
- CODEX_IMAGE
- GEMINI_IMAGE
- GEMINI_AUTH_PATH
- GOOGLE_CLOUD_PROJECT
- DEFAULT_TIMEOUT
- LOG_LEVEL
- MCP_PROFILES_DIR
- MCP_LOG_LEVEL
- PYTHONPATH
- CODEX_AUTH_PATH
- MCP_COMPLEXITY_ROUTING
- MCP_COMPLEXITY_WORD_HIGH
- MCP_COMPLEXITY_WORD_MEDIUM
- MCP_COMPLEXITY_PREFER_THRESHOLD
- USE_SEMANTIC_ROUTING
- SEMANTIC_ROUTING_ALPHA
- SEMANTIC_MODEL_NAME
- SEMANTIC_CACHE_ENABLED
- SEMANTIC_BATCH_SIZE
- ROUTING_DEBUG
- BENCHMARK_MODE
Security Notes
The server orchestrates running AI-generated code in Docker containers, which inherently carries risk. It mitigates this with per-run containers, explicit 'readonly', 'workspace_write', and 'full_access' security modes, and read-only mounting of sensitive authentication files (`auth.json`). The 'full_access' mode, while documented, is inherently dangerous and allows the AI to make arbitrary changes.
Similar Servers
claude-prompts-mcp
Enhances AI assistant behavior through structured prompt management, multi-step chains, quality gates, and autonomous verification loops, primarily for development tasks.
wanaku
Centralized routing and management of AI agent access to diverse tools and resources via the Model Context Protocol (MCP).
mcp-ai-agent-guidelines
A comprehensive AI agent development framework focused on structured design, prompt engineering, code analysis, and agent-to-agent orchestration for developers.
dotprompts
A personal prompt management system exposed as a Model Context Protocol (MCP) server, enabling AI agents to access, create, update, and delete user-defined prompts.