Jacky CLI

AI CLI, Automate.

Designer / Author: Maturi Jaswanth Sai Madhu Mohan

Built on Hermes Agent by Nous Research. Jacky CLI is a personalized, distinct distribution of Nous Research's MIT-licensed Hermes Agent — full credit and thanks to the Nous Research team for the original agent, its tool-calling architecture, and its self-improving skill system. This fork keeps that foundation and adds: a bundled offensive-security / bug-bounty hunt-loop methodology (skills/, METHODOLOGY.md), a one-command setup.sh bootstrap, and CLI ergonomics tuned around dual local + cloud model use. If you're looking for the upstream project, it's at github.com/NousResearch/hermes-agent.

Quick Start

git clone https://github.com/jaswanthsai1/jacky-cli.git
cd jacky-cli
./setup.sh
jacky

setup.sh creates a virtual environment, installs Jacky CLI into it, copies .env.example → .env, links the jacky command onto your PATH, and — before declaring success — actually runs jacky --help to prove the install works.

Windows (native, PowerShell):

git clone https://github.com/jaswanthsai1/jacky-cli.git
cd jacky-cli
powershell -ExecutionPolicy ByPass -File scripts\install.ps1

See website/docs/user-guide/windows-native.md for the native Windows feature matrix.

📖 Full documentation →  |  🎯 Hunt-loop methodology →

What Jacky can do

Local model (Ollama) vs. cloud provider setup

Jacky supports both, and switching between them is a one-line config change.

Local, zero API cost, fully offline:

curl -fsSL https://ollama.com/install.sh | sh   # install Ollama
ollama pull qwen3:8b                            # or any tool-calling-capable model
jacky model                                     # pick Ollama + the model you pulled

CPU-only works but is slower — see website/docs/guides/local-ollama-setup.md for hardware guidance, model recommendations, and the timeout tuning needed for slow CPU inference.

Cloud, any OpenAI-compatible provider:

cp .env.example .env   # done for you by setup.sh
# edit .env: add the API key for OpenRouter, OpenAI, Anthropic, Google AI
# Studio, z.ai, Kimi, MiniMax, Hugging Face, or your own OpenAI-compatible
# endpoint — see .env.example for the full list
jacky model            # pick your provider and model

Full provider reference: website/docs/integrations/providers.md.

Getting Started

jacky              # Interactive CLI — start a conversation
jacky model        # Choose your LLM provider and model
jacky tools        # Configure which tools are enabled
jacky config set   # Set individual config values
jacky gateway      # Start the messaging gateway (Telegram, Discord, etc.)
jacky setup        # Run the full setup wizard (configures everything at once)
jacky update       # Update to the latest version
jacky doctor       # Diagnose any issues

📖 Full documentation →  |  🎯 Hunt-loop methodology →

CLI vs Messaging Quick Reference

Jacky has two entry points: start the terminal UI with jacky, or run the gateway and talk to it from Telegram, Discord, Slack, WhatsApp, Signal, or Email. Once you're in a conversation, many slash commands are shared across both interfaces.

For the full command lists, see website/docs/user-guide/cli.md and website/docs/user-guide/messaging/.

Documentation

Source docs live under website/docs/:

Contributing

We welcome contributions! See CONTRIBUTING.md for development setup, code style, and PR process.

git clone https://github.com/jaswanthsai1/jacky-cli.git
cd jacky-cli
./setup.sh
.venv/bin/pip install -e ".[all,dev]"
scripts/run_tests.sh

Community

• 🐛 Issues
• 📚 Skills Hub (agentskills.io)
• 🔌 computer-use-linux — Linux desktop-control MCP server for Jacky and other MCP hosts, with AT-SPI accessibility trees, Wayland/X11 input, screenshots, and compositor window targeting.

License

MIT — see LICENSE.

Jacky CLI is a fork of Hermes Agent, © Nous Research, used and modified under the MIT License. Jacky-specific additions © Maturi Jaswanth Sai Madhu Mohan. See LICENSE for the full dual attribution.