JD.AI

An AI-powered terminal assistant and multi-channel platform built on Microsoft Semantic Kernel. Connect to 15 AI providers — from cloud APIs to fully local models — and get a complete coding agent in your terminal.

Get started · View on GitHub · NuGet package

JD.AI terminal session

Install in one command

dotnet tool install --global JD.AI
jdai

JD.AI auto-detects available providers. No manual configuration needed.

What you can do

Task Example prompt
Explore codebases what does this project do?
Fix bugs the login fails after timeout, fix it and verify with tests
Refactor code refactor the auth module to use async/await
Write tests write unit tests for the Calculator.Divide method
Create commits commit my changes with a descriptive message
Spawn subagents use an explore agent to find how caching works
Orchestrate teams use a debate team to discuss microservices vs monolith
Search the web what are the latest features in .NET 10?

Chat with streaming responses

Features at a glance

Feature Description
Multi-provider 15 AI providers including Claude Code, Copilot, Ollama, OpenAI, Anthropic, Azure, Gemini, Bedrock, Foundry Local, and more
35+ slash commands Model switching, sessions, context, workflows, defaults
17 tool categories Files, search, shell, git, web, memory, subagents, think, tasks, code execution, clipboard, questions, diff/patch, batch edit, usage tracking
Tool loadouts Curated tool bundles — expose only the tools each agent needs; 5 built-ins (minimal, developer, research, devops, full)
5 subagent types Explore, task, plan, review, general-purpose
4 orchestration strategies Sequential, fan-out, supervisor, debate team coordination
6 channel adapters Discord, Signal, Slack, Telegram, Web, OpenClaw
Session persistence Save, load, export conversations across sessions
MCP integration Connect external tool servers via Model Context Protocol
Workflows Composable multi-step automation with /workflow commands
Observability OpenTelemetry tracing, metrics, health checks
Git checkpointing Safe rollback with stash/directory/commit strategies
Skills & plugins Native managed skills/plugins/hooks lifecycle
Auto-update Check and apply updates via NuGet from your terminal

Subagent execution and team orchestration

Documentation

User Guide — Using JD.AI

For anyone using jdai as a tool. Installation, provider setup, commands, workflows, and best practices.

Developer Guide — Extending & integrating

For developers building on JD.AI. Architecture, custom tools, providers, plugins, and the gateway API.

Specifications (UPSS) — Living product specs

The Unified Product Specification System — 18 machine-verifiable spec types covering product intent through operational governance.

Operations — Deploy, monitor, govern

For ops teams and administrators. Deployment, observability, security, and enterprise governance.

Reference — Quick lookup

Complete reference material for CLI flags, commands, tools, providers, and configuration.