We're trying to fix a context-sync problem between frontend and backend development environments, especially when using AI coding assistants that need shared context to stay in sync. * 🧩 The Root Problem When the frontend and backend live in separate IDE sessions: The AI assistant in the frontend doesn’t know about backend updates (new endpoints, changed schemas, etc.). You have to manually describe the changes (via API.md or OpenAPI export). Copying/syncing specs between folders and contexts becomes tedious and fragile. The problem is not code synchronization itself (Git already does that), but AI context synchronization between two projects that are logically linked. Most tools solve spec synchronization (OpenAPI <-> frontend/backend), but none natively integrate with AI coding assistants or automatically update contextual knowledge (like your `API.md` file feeding AI prompts). That's the purpose of Conciliate — a local schema bridge between multiple IDE AI sessions. It aims to extend existing OpenAPI generators with: - IDE-aware context streaming for AI models. - Language-agnostic plugin hooks. - Lightweight auto documentation. In short tools like Apidog, DeepDocs, and NSwag already cover the file and schema sync aspects, but the “AI context sync” layer remains an open gap. We went straight for an efficient, Python-based MVP using FastAPI + MCP, designed to work via three access modes: 1️⃣ MCP server (for AI assistants / extensions), 2️⃣ IDE extension 3️⃣ CLI **Goal: Keep frontend and backend (and their AI assistants) automatically synchronized by maintaining a live, shared API context that updates whenever backend APIs change. A lightweight local service that watches the backend, auto-extracts the API spec, summarizes changes, and exposes this data to AI assistants and IDEs — no manual `API.md` copying.