--- mw_bundle: 1 id: EJbI7cto title: "AI tool integration tour" url: https://memory.wiki/b/EJbI7cto document_count: 4 updated: 2026-05-14T19:20:01.270Z analysis_generated_at: 2026-05-14T19:20:01.270Z source: "memory.wiki" --- # AI tool integration tour > Claude Code, Codex (AGENTS.md), MCP server, Cursor — every wiring path the demo hub supports. **Intent:** Curated bundle: AI tool integration tour ## Summary This collection provides a comprehensive guide for integrating mdfy knowledge management system across four major AI development tools. The documents demonstrate both simple URL-based context injection and advanced MCP server setup, covering configuration patterns, optimization techniques, and common pitfalls for each platform. ## Themes - Cross-platform AI tool integration - Progressive complexity from URL-based to MCP protocol - Context scoping strategies - Configuration management patterns ## Cross-document insights - The documentation reveals a deliberate architecture progression from simple URL fetching to full bidirectional MCP integration, suggesting mdfy is designed for both quick adoption and advanced use cases - Each AI tool has distinct configuration contracts (CLAUDE.md, .cursor/rules, AGENTS.md) but they all converge on the same URL-based context pattern, indicating potential for standardization - The consistent mention of token optimization across tools suggests context window management is a universal constraint that mdfy's compact flags specifically address - MCP represents a paradigm shift from read-only context to active knowledge management, but only one document covers it fully, indicating this may be an emerging capability ## Key takeaways - mdfy provides both simple URL-based context injection and advanced MCP server integration to suit different complexity needs across AI development tools - Each AI tool requires platform-specific configuration files but follows consistent patterns for context loading and token optimization - Session restart requirements and caching behaviors vary significantly between tools and represent the primary integration gotchas to watch for ## Open questions / gaps - No comparison matrix showing which features are available across different integration approaches - Missing troubleshooting section for common authentication and URL formatting errors - No performance benchmarks or context window usage comparisons between different optimization flags ## Notable connections - **doc:iok3gZo6** ↔ **doc:PrxR1Uj4** — Simple URL integration from Claude Code doc can be upgraded to full MCP server capabilities shown in the MCP setup guide - **doc:TEuiwop8** ↔ **doc:y4sEEo1R** — Both demonstrate similar URL-based patterns but reveal platform-specific caching behaviors that affect configuration management - **doc:iok3gZo6** ↔ **doc:y4sEEo1R** — Both emphasize session restart requirements but Claude Code treats CLAUDE.md as a contract while Codex caches AGENTS.md more aggressively ## Concepts (this bundle) - **Model Context Protocol** - **URL-based Context** - **Context Scoping** - **Session Persistence** - **Hot-reload Limitations** ## Concept relations - **Model Context Protocol** ↔ **URL-based Context** — extends beyond ## Documents ### 1. [Wiring mdfy into Cursor](https://memory.wiki/TEuiwop8) `.cursor/rules/mdfy.mdc`: *sections:* How Cursor uses this: On every new chat or composer session, Cursor evaluates files matched by the rule and includes them as context. When mdfy URL is in the rule body, Cursor fetche | What we observed in practice: First few messages: Cursor doesn't always fetch — it relies on its own RAG over the local repo. Fine.; Once the user types something topic-specific ("how did we decide the pricing tier?"), Cursor fetches the bundle URL and the answer comes back cited to the right; Cursor's [doc:N] citations don't propagate the way Claude Chat's do — Cursor inlines snippets but doesn't link back. Acceptable for a code-editor surface. | … ### 2. [Wiring mdfy into Claude Code](https://memory.wiki/iok3gZo6) Tested against Claude Code v1.x. Single command, single line in CLAUDE.md, persistent across sessions. *sections:* What you put in CLAUDE.md: Drop this somewhere near the top of the file (CLAUDE.md is loaded into every Claude Code session as system context): | Context: When you need background on this project, fetch: | Why this works: Claude Code respects CLAUDE.md. It's not a hint; it's a contract. Anything in CLAUDE.md is in the session's persistent system prompt.; mdfy hubs are URL-shaped. No SDK, no auth, no streaming setup — just a GET that returns markdown plus an inlined graphdata JSON block.; Token cost is tunable. Add ?compact to the URL to strip low-density sections. Most users save 30-40% of the context window this way. | … ### 3. [Wiring mdfy into Codex via AGENTS.md](https://memory.wiki/y4sEEo1R) Codex CLI + AGENTS.md. Same shape as the Claude Code recipe but with caveats specific to Codex's caching. *sections:* What you put in AGENTS.md: Drop this near the top: | Context: Load https://mdfy.app/hub/ before answering project | Project bundle: https://mdfy.app/b/?compact | Why two scopes (hub + bundle): The hub gives the AI breadth: who you are, what you're building, what your prior work shows. The bundle gives it depth: the specific docs related to the current | What's different from Claude Code: Codex caches AGENTS.md at the start of a session. If you edit AGENTS.md mid-session, Codex doesn't see the change until the next session. This caught me twice o | … ### 4. [mdfy MCP server: setup in 4 lines](https://memory.wiki/PrxR1Uj4) For Claude Desktop, Cursor, and any other MCP-enabled tool. *sections:* What MCP gives us: MCP (Model Context Protocol) is Anthropic's open standard for letting AI tools call external services. The mdfy MCP server exposes seven tools — read, search, c | Setup: Add this to your MCP config (Claude Desktop: ~/Library/Application Support/Claude/mcpconfig.json; Cursor: .cursor/mcp.json in the project): | Authentication: The MCP server reads MDFYEDITTOKEN and MDFYUSEREMAIL from the environment. Add them to the MCP config like this: | The tools exposed: | Tool | Purpose | | What changes in your workflow: Before MCP: you ask Claude "summarise the launch plan", and Claude says "I'd need to read the plan — paste it for me." After MCP: Claude calls mdfyread with the _Digest view — follow any link above to fetch that doc's full markdown. Add `?full=1` to this URL for the concatenated payload._