glm-delegator

LLM Delegator

Multi-provider LLM expert subagents for Claude Code — A fork of claude-delegator supporting any LLM backend

LLM expert subagents for Claude Code. Five specialists that can analyze AND implement—architecture, security, code review, and more.

Supports: Anthropic Claude, OpenAI GPT, GLM-4.7, GLM-5, Ollama, Groq, DeepInfra, and any OpenAI/Anthropic-compatible API.

What is LLM Delegator?

Claude gains a team of LLM specialists via MCP. Each expert has a distinct specialty and can advise OR implement.

What You Get	Why It Matters
5 domain experts	Right specialist for each problem type
Dual mode	Experts can analyze (read-only) or implement (write)
Multi-provider	Use Claude, GPT-4, GLM (4.7/5), Ollama, or any compatible API
Auto-routing	Claude detects when to delegate based on your request
Prompt Enhancement	LLM-based prompt improvement before expert delegation
Synthesized responses	Claude interprets LLM output, never raw passthrough
Multilingual	Code Review supports EN/FR/CN (中文)

The Experts

Expert	What They Do	Example Triggers
Architect	System design, tradeoffs, complex debugging	"How should I structure this?" / "What are the tradeoffs?"
Plan Reviewer	Validate plans before you start	"Review this migration plan" / "Is this approach sound?"
Scope Analyst	Catch ambiguities early	"What am I missing?" / "Clarify the scope"
Code Reviewer	Find bugs, improve quality (EN/FR/CN)	"Review this PR" / "What's wrong with this?"
Security Analyst	Vulnerabilities, threat modeling	"Is this secure?" / "Harden this endpoint"

Differences from claude-delegator

Feature	claude-delegator	llm-delegator
Backend	Codex GPT-5.2 only	Any LLM provider
Configuration	CLI args only	CLI args (same model)
Providers	Single	Multi-provider (OpenAI, Anthropic, Ollama, etc.)
Code Review	English only	EN/FR/CN multilingual
Security	OWASP	OWASP + Chinese MLPS standards
Prompt Enhancement	None	LLM-based auto-enhancement
Routing	None	Intelligent task routing
Workflow	None	State machine for automation
License	MIT	MIT

Install

Prerequisites

• Python 3.8+ for the MCP server
• API Key for your chosen provider (Anthropic, OpenAI, Z.AI, etc.)
• httpx library - pip install -r requirements.txt

Step 1: Install Dependencies

cd glm-delegator
pip install -r requirements.txt

Step 2: Configure API Key

Set your API key as an environment variable:

# Anthropic Claude
export ANTHROPIC_API_KEY="sk-ant-..."

# OpenAI
export OPENAI_API_KEY="sk-..."

# GLM via Z.AI
export GLM_API_KEY="your_z_ai_api_key_here"

# Groq
export GROQ_API_KEY="gsk_..."

For persistent configuration, add to your ~/.bashrc or ~/.zshrc:

echo 'export ANTHROPIC_API_KEY="sk-ant-..."' >> ~/.bashrc
source ~/.bashrc

Step 3: Register MCP Server

Add to ~/.claude.json (or ~/.claude/settings.json):

Using Anthropic Claude

{
  "mcpServers": {
    "claude-experts": {
      "type": "stdio",
      "command": "python3",
      "args": [
        "/path/to/glm-delegator/glm_mcp_server.py",
        "--provider", "anthropic-compatible",
        "--base-url", "https://api.anthropic.com/v1",
        "--api-key", "$ANTHROPIC_API_KEY",
        "--model", "claude-sonnet-4-20250514"
      ]
    }
  }
}

Using OpenAI GPT

{
  "mcpServers": {
    "openai-experts": {
      "type": "stdio",
      "command": "python3",
      "args": [
        "/path/to/glm-delegator/glm_mcp_server.py",
        "--provider", "openai-compatible",
        "--base-url", "https://api.openai.com/v1",
        "--api-key", "$OPENAI_API_KEY",
        "--model", "gpt-4o"
      ]
    }
  }
}

Using Ollama (Local)

{
  "mcpServers": {
    "ollama-experts": {
      "type": "stdio",
      "command": "python3",
      "args": [
        "/path/to/glm-delegator/glm_mcp_server.py",
        "--provider", "openai-compatible",
        "--base-url", "http://localhost:11434/v1",
        "--model", "llama3.1"
      ]
    }
  }
}

Using GLM via Z.AI

{
  "mcpServers": {
    "glm-experts": {
      "type": "stdio",
      "command": "python3",
      "args": [
        "/path/to/glm-delegator/glm_mcp_server.py",
        "--provider", "anthropic-compatible",
        "--base-url", "https://api.z.ai/api/anthropic",
        "--api-key", "$GLM_API_KEY",
        "--model", "glm-5"
      ]
    }
  }
}