Your Multi-Agent Command Center

Chakravyuh AI orchestrates specialized AI agents — coder, researcher, planner, QA, security — into a coordinated mesh that runs on your hardware. Terminal-first, Ollama-powered, open-source.

$ cv run
⚔ Starting Chakravyuh AI...
✓ Ollama detected (llama3.1:8b, qwen2.5:7b, mistral:7b)
✓ 10 agents registered
✓ Control Agent online
Dashboard: http://127.0.0.1:3001/

Why Chakravyuh?

A new approach to multi-agent AI — designed for local-first, developer-controlled agent orchestration

🤖

Control Agent

An intelligent orchestrator that decomposes complex tasks, assigns subtasks to specialized agents, and monitors execution in real-time.

🌟

Terminal-First

Rich CLI + Textual TUI. Full control from the command line: cv run, cv chat, cv status, cv tui.

Model Router

Three modes: local (always Ollama), hybrid (Ollama + cloud), cloud (always cloud). Switch with one env var.

📖

Tiered Memory

Working, episodic, semantic, procedural memory with SQLite or ChromaDB backends. Swap backends with an env variable.

💻

MCP Tools

Filesystem, web fetch/search, GitHub, Docker, terminal — all available to agents via a uniform tool interface.

🌐

Web Dashboard

React + Vite light-themed UI for visual agent management, task submission, and model management.

Agent Mesh

10 specialized agents collaborating through the Control Agent

C
Coordinator
Task routing & delegation
llama3.1:8b
P
Planner
Goal decomposition & planning
llama3.1:8b
C
Coder
Write, review & refactor code
qwen2.5:7b
R
Researcher
Information gathering & analysis
llama3.1:8b
B
Browser
Web automation & interaction
mistral:7b
Q
QA
Testing, validation & QA
gemma2:9b
M
Memory Agent
Memory management & retrieval
nomic-embed-text
S
Security
Security analysis & threat detection
llama3.1:8b
G
GitHub Agent
Repository management & automation
qwen2.5:7b
D
Deployment
Build, deploy & infrastructure
phi3:14b

One-Command Control

Everything is a cv command away

cv setup — Auto-detect Ollama, pull models, create .env cv run — Start the API server + open dashboard cv tui — Launch the Textual terminal dashboard cv status — System health overview cv models --list — List installed Ollama models cv models --pull llama3.2:3b — Pull a new model cv chat -m llama3.1:8b — Interactive chat session cv task "Build a CLI tool" — Submit to agent mesh cv agents — List all registered agents

How to Collaborate

Chakravyuh is open-source and community-driven. Everyone is welcome.

📜 Submit an Issue

Found a bug or have a feature request? Open an issue on GitHub with the appropriate template.

Open Issue

🛠 Contribute Code

Fork the repo, create a feature branch, and submit a PR. All contributions are reviewed and appreciated.

View PRs

💬 Join the Discussion

Questions, ideas, and discussions happen on GitHub Discussions. Introduce yourself and share your thoughts.

Start Discussion

📖 Read the Docs

Comprehensive documentation covering setup, architecture, agents, MCP, models, and governance.

Read Docs

🌐 Governance

BDFL + Core Maintainer model with clear roles, decision-making process, and contributor progression path.

View Governance

⭐ Star on GitHub

Show your support by starring the repository. It helps others discover the project.

Star Now