dev-mcp-server

> AI that understands your codebase, not just the internet.

Inspired by *"How I Built an MCP Server That Made Developers Faster and Work Easier"* — a full implementation of the Model Context Platform concept: instead of generic AI answers, every response is grounded in your actual code, error logs, API behavior, and bug history.

---

Every team has this invisible tax:

• Debugging code you didn't write, with zero context
• Explaining things that are already written *somewhere*
• Digging through 10 files to understand one API
• Answering the same question for the third time this week

The root cause isn't bad code. It's a context problem — knowledge scattered across services, logs, configs, and people's heads.

---

Before answering any question, the AI looks up your actual system. It knows:

• Your data models and DTOs
• Your naming conventions and code patterns
• Your most common bugs and how you fixed them
• Your API behaviour — including weird edge cases
• How your modules connect to each other

---

| Query | Endpoint | Example | | ------------------------------------ | ------------------------ | ---------------------------------------------------- | | 🐛 Why is this failing? | POST /api/query/debug | "Why is ClassCastException thrown in UserService?" | | 🔍 Where is this used? | POST /api/query/usage | "Where is getUserById called?" | | 💥 If I change this, what breaks? | POST /api/query/impact | "If I change the User model, what breaks?" |

---

# In your project root (where your .env lives):
npx dev-mcp-server ingest ./src
npx dev-mcp-server query "Why is getUserById throwing?"
npx dev-mcp-server que