focal-mcp
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@focal-mcpfetch the current AI behavior rules"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
FOCAL MCP Server
FOCAL MCP (Framework for Orchestrated Central AI ruLes) is a local MCP server that centralizes AI behavior rules and applies them across tools like Codex, Claude Code, and Cursor.
Quick Start
Prerequisites
uvinstalled (required foruvx)
Run the server (recommended)
uvx --from focal-mcp-server focal-mcp-serverOpen the Web UI:
http://127.0.0.1:8765MCP endpoint:
http://127.0.0.1:8765/mcpWebSocket notifications:
ws://127.0.0.1:8765/mcp/wsRelated MCP server: standards-mcp
Configure Your MCP Client
FOCAL MCP uses streamable HTTP, so you start the server yourself and point clients to the URL above.
Codex (example)
Add the following to ~/.codex/config.toml:
[mcp_servers.focal]
enabled = true
url = "http://127.0.0.1:8765/mcp"Restart Codex after saving.
Edit Rules (Live)
Rules live outside the repo in:
~/.focal_mcp/workspace/
core/
system.md
style.md
safety.md
tool_policy.md
agents/Use the Web UI to edit files. Updates are applied immediately.
How Rules Are Applied
FOCAL MCP exposes a tool named focal_rules. Agents should call it before every response to fetch the latest rules.
If rules do not update immediately:
Ensure the client is connected to the MCP server.
Ensure the client calls
focal_rulesbefore responding.Reconnect the client if needed.
Project Prompt (Codex)
The repository-level prompt is stored in:
agents.mdRoadmap
Project-specific rule management for multi-repo workflows
UI/UX improvements for rule editing and visibility
This server cannot be deployed
Maintenance
Related MCP Connectors
The system of record for AI agent authority: playbooks, routed policy questions, reusable rules.
Path-scoped team memories, rules and skills for Claude Code, Cursor, Codex and other MCP clients.
Research-backed linting + generation for agent context files (CLAUDE.md, AGENTS.md, Cursor rules).
Persistent cross-session memory shared by Codex, Claude Code, ChatGPT, and other AI agents.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceProvides AI agents with queryable, version-controlled project rules and coding standards. Enables validation, rule-based guidance, and task summaries to keep AI work aligned with your project's conventions without repeating context.2MIT
- AlicenseNot gradedqualityCmaintenanceTransforms static coding standards into a queryable live data store for AI agents, delivering task-specific rules and fix guidance on demand. This optimizes context window usage through progressive disclosure, ensuring agents apply relevant governance without loading massive documentation.2MIT
- AlicenseCqualityCmaintenanceCaptures code modifications via AST analysis, generates reusable rules, and injects them into AI agent context with version audit and token-budgeted rule retrieval.2811 npmApache 2.0

gnt MCP Serverofficial
AlicenseNot gradedqualityAmaintenanceEnables AI agents to query live, human-approved rules before taking actions, ensuring compliance and reducing errors.30Apache 2.0