🤖 DevMind

DevMind is a context-aware AI assistant for developers that runs in your terminal. It uses LLMs (like GPT-4 or Code Llama) to explain code, suggest fixes, analyze logs, and generate commands based on your working environment.

Think of it as copilot, but for your terminal – with deep integration into your codebase, shell, and dev tools.

List of Commands

Command	Description
devmind chat	Chat with your project (With memory)
devmind explain	Explain a source code file
devmind summarize	Summarize code, logs, or text
devmind generate	Generate files based on descriptions

🚀 Features

• 🧑‍💻 Chat with devmind ask any questions
• 🧠 Memory for context-aware conversations
• 📜 Explain source code files (.go, .py, .js...)
• 📖 Summarize code, logs, or text
• 🛠️ Generate shell commands, Dockerfiles, Makefiles
• 🔌 Pluggable backends: Ollama, (in Development): OpenAI

📦 Example Usage

$ devmind chat what did we discussed in the last chat?
"Last time, we discussed the new feature for user authentication..."

$ devmind chat what is the purpose of this function?
"This function handles user login by validating credentials and generating a token..."

$ devmind generate create a Dockerfile for a Go web server
"FROM golang:1.21...."

$ devmind explain --file main.go --line 10-15
"This function is responsible for handling user login requests. It takes the username and password as input, validates them, and returns a token if successful..."

$ devmind summarize --file main.go
"This Go program initializes a context with signal handling for graceful shutdown. It sets up..."

📁 Project Structure

devmind/
├───bin
├───cmd                 # CLI commands
│   └───ui              # UI elements (Spinner)
├───config              # code for config
├───internal
│   ├───agent           # Agents for LLM
│   │   ├───ollama
│   │   ├───openai
│   │   └───types
│   ├───consumer        # LLM consumer
│   ├───handlers        # handlers for different commands
│   ├───logger          # logging and error handling
│   ├───memory          # code for chat memory
│   │   └───chat
│   └───utils           # Shell wrappers, file utils
└── main.go

⚙️ Requirements

Component	Required For
Go 1.24+	CLI Tool
Ollama Endpoint	endpoint where Ollama is hosted

💡 Features

Capability	Description
✅ Chat Continuity	Context-aware conversations stored in memory
✅ Response Streaming	Stream response as it is generated
✅ Responsive UX	Spinner to show progress and status updates
✅ Code summarization	Explain Go/Python/JS files and code sections
✅ Log diagnosis	Parse and interpret crash/error logs
✅ Script generation	Generate shell commands, Dockerfiles, Makefiles
✅ File summarization	Summarize code, logs, or text files
✅ Multi-backend LLMs	Use Ollama, (OpenAI, Claude planned)
🔒 Privacy-first	Full local-only mode available

🧠 Memory System

Currently, DevMind has a basic memory system that stores chat history in a file. This allows for context-aware conversations and helps the AI remember previous interactions.

Future plan: Summarize and store chat history for better context in conversations.

• After each chat or command, DevMind spawns a background summarizer.
• Summaries are stored inside summaries/context.txt.
• Future prompts inject recent memory to improve response continuity.

✅ Non-blocking summarization ✅ Fast startup and exit ✅ Future upgrade path for semantic search memory

🧠 Supported LLMs

• ✅ Local Ollama models (CodeLlama, Mistral, Phi)
• ✅ OpenAI (gpt-3.5, gpt-4) (planned)
• ✅ Anthropic (Claude) (planned)

🧑‍💻 Author

Made with Go, LLMs, and caffeine by @vinit-chauhan

📄 License

MIT © Vinit Chauhan