model-council-mcp-codex
by tsarihan
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| GROK_CLI | No | Set to 'true' to add subscription-backed Grok members via local grok CLI | false |
| CODEX_CLI | No | Set to 'true' to add subscription-backed ChatGPT member via local codex exec CLI | false |
| MAX_FILES | No | Max number of attached files | 32 |
| CLAUDE_CLI | No | Set to 'true' to add subscription-backed Claude members via the local claude CLI | false |
| MAX_TOKENS | No | Max output tokens per completion | 32768 |
| JUDGE_MODEL | No | Judge model ID or 'auto' | auto |
| MAX_FILE_KB | No | Per-file cap in KB | 512 |
| XAI_API_KEY | No | Enables X.AI Grok models | |
| MAX_TOTAL_KB | No | Total across all attached files in KB | 1536 |
| VLLM_SERVERS | No | Comma-separated list of name:host:port for vLLM servers | |
| GROK_CLI_PATH | No | Path to the grok binary | grok |
| RESPONSE_MODE | No | Response mode: individual, categorized, deconflicted, pooled, dialectic | categorized |
| CODEX_CLI_PATH | No | Path to the codex binary | codex |
| COUNCIL_MODELS | No | Comma-separated model IDs | |
| MAX_CONTEXT_KB | No | Inline context string cap in KB | 1024 |
| OLLAMA_ADDRESS | No | Base URL of your Ollama server | http://localhost:11434 |
| OPENAI_API_KEY | No | Enables OpenAI models | |
| SGLANG_SERVERS | No | Comma-separated list of name:host:port for SGLang servers | |
| TRTLLM_SERVERS | No | Comma-separated list of name:host:port for TRT-LLM servers | |
| CLAUDE_CLI_PATH | No | Path to the claude binary | claude |
| GROK_CLI_MODELS | No | Model names for the Grok CLI member | grok-4.5 |
| MAX_QUESTION_KB | No | Question string cap in KB | 256 |
| CODEX_CLI_MODELS | No | Model names for the Codex member | default |
| ANTHROPIC_API_KEY | No | Enables Anthropic Claude models | |
| CLAUDE_CLI_MODELS | No | Model aliases for the Claude CLI member | opus,sonnet |
| CLOUD_CONCURRENCY | No | Max simultaneous requests to cloud members | 3 |
| LOCAL_CONCURRENCY | No | Max simultaneous requests to local models; 0 = unlimited | 1 |
| COMPLETION_RETRIES | No | Attempts per completion before giving up | 3 |
| DECONFLICT_VERBOSE | No | If 'true', deconflicted results include per-round detail by default | false |
| REQUEST_TIMEOUT_MS | No | Per-completion wall-clock timeout in ms for text-only calls | 300000 |
| MODEL_COUNCIL_STATE | No | Path to persistent state file (overrides ~/.config/model-council/state.json) | |
| MAX_DECONFLICT_ROUNDS | No | Max deconfliction iterations | 3 |
| REPO_REQUEST_TIMEOUT_MS | No | Per-completion timeout in ms when full_repo_access is set | 600000 |
| CLAUDE_CLI_OLLAMA_MODELS | No | Comma-separated Ollama model names to run through the claude CLI harness | |
| CLAUDE_CLI_OLLAMA_ADDRESS | No | Ollama address for the harness member, if different from OLLAMA_ADDRESS | |
| GROK_CLI_UNSAFE_ACCEPT_RCE | No | Set to 'true' to re-enable Grok CLI members (unsafe, testing only) | false |
Capabilities
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
No tools | |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/tsarihan/model-council-mcp-codex'
If you have feedback or need assistance with the MCP directory API, please join our Discord server