{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreifnpg3jo3thw6uxaiwykbxwsgqzrp4ituglcrdywv5t7eejs2gbty",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mopejnaxqr22"
  },
  "path": "/t/coding-tools-mcp-turning-chatgpt-into-a-real-coding-agent/1384178#post_1",
  "publishedAt": "2026-06-20T06:03:35.000Z",
  "site": "https://community.openai.com",
  "textContent": "Hi everyone,\n\nI’ve been experimenting with MCP-based workflows recently, and one thing kept bothering me:\n\nChatGPT is incredibly capable at writing code, but it normally can’t interact with a real project.\n\nIt can’t inspect a repository, search files, run commands, check Git status, or apply changes directly. Most of the workflow still involves copying and pasting context back and forth.\n\nTo address this, I built **Coding Tools MCP** , an MCP server that exposes practical coding operations to ChatGPT and other MCP-compatible clients.\n\n## What it provides\n\nCoding Tools MCP exposes a set of tools commonly needed during software development:\n\n  * Read files and browse directories\n\n  * Search across a codebase\n\n  * Apply patches to files\n\n  * Execute terminal commands\n\n  * Interact with long-running processes\n\n  * Inspect Git repositories\n\n  * Request permission escalation when needed\n\n  * View local images\n\n\n\n\nThe goal is simple:\n\n**Give AI assistants direct, controlled access to a real development workspace.**\n\n## Example workflow\n\nInstead of:\n\n> “Here’s a file. Can you review it?”\n\nYou can ask:\n\n> “Analyze this repository and explain its architecture.”\n\nInstead of:\n\n> “Here’s the error log.”\n\nYou can ask:\n\n> “Find the root cause of this bug and propose a fix.”\n\nInstead of:\n\n> “Generate a patch.”\n\nYou can ask:\n\n> “Implement this feature and update the relevant files.”\n\nThe model can then use the MCP tools to inspect the project, search code, run commands, and modify files as needed.\n\n## Security\n\nThe project was designed around workspace confinement and explicit permissions.\n\nIt supports:\n\n  * Workspace-scoped file access\n\n  * Permission requests for sensitive operations\n\n  * Runtime policy controls\n\n  * Git-aware tooling\n\n  * Process lifecycle management\n\n\n\n\nThe intention is to make powerful coding capabilities available while keeping operations transparent and auditable.\n\n## Why I built it\n\nI wanted something that felt closer to a coding agent than a traditional chatbot, while remaining compatible with the MCP ecosystem and easy to self-host.\n\nWith ChatGPT Apps and MCP support becoming more widely available, it felt like the right time to build a dedicated coding-focused MCP server.\n\n## Project\n\nGitHub: xyTom/coding-tools-mcp\n\nI’d love to hear feedback, feature requests, or ideas from others building MCP-powered developer tools.",
  "title": "Coding Tools MCP: Turning ChatGPT into a Real Coding Agent"
}