English | 简体中文 | 日本語 | 한국어 | Español | Português | العربية | Deutsch | Français | Türkçe | Русский

Mysti - Your AI Coding Team Working Together

Your AI Coding team for VSCode
12 AI providers — Claude Code, Codex, Gemini, Copilot, Cline, Cursor, OpenClaw, Manus, OpenCode, Qwen Code, Ollama & LocalAI — working solo or in teams
Wisdom of the crowd where the collective intelligence of several agents outperforms a single one.

Providers • Brainstorm • Features • Quick Start • Config • Docs

---

What's New in v0.4.0

4 New Providers (12 Total)

Mysti now supports 12 AI providers — added OpenCode, Qwen Code, Ollama, and LocalAI alongside Claude Code, Codex, Gemini, GitHub Copilot, Cline, Cursor, OpenClaw, and Manus. Run local models with Ollama/LocalAI or use cloud providers like OpenCode and Qwen Code. Each provider has its own authentic logo in the UI.

Qwen Code

Alibaba's AI coding CLI with deep reasoning capabilities. Uses the same streaming protocol as Claude Code for seamless integration. Supports Qwen3 Coder models with plan, auto-edit, and yolo approval modes.

OpenCode

Multi-backend coding agent supporting Anthropic, OpenAI, Google, and Groq through a single CLI. Uses your configured default model — no lock-in to specific providers.

Local AI Support

Run AI models locally with Ollama and LocalAI — no cloud subscription needed. Full privacy, zero latency, complete control over your models.

Stability & Testing

18 stability fixes across brainstorm mode and @-mention tagging — silence-based timeouts, auth pre-checks, convergence guards, mention limits, and retry cleanup. Plus 360 automated tests via vitest ensuring reliability across all features.

Bug Fixes

• Windows: Fixed spawn EINVAL error with auto shell mode (#14)
• Codex: Brainstorm now respects mysti.codexPath setting (#26)

---

Install in Seconds

From VS Code: Press Ctrl+P (Cmd+P on Mac), then paste:

ext install DeepMyst.mysti

Or install from the VS Code Marketplace

---

Choose Your AI

Mysti works with the AI coding tools you already have. No extra subscriptions needed.

Provider	Best For
Claude Code	Deep reasoning, complex refactoring, thorough analysis
Codex	Quick iterations, familiar OpenAI style
Gemini	Fast responses, Google ecosystem integration
GitHub Copilot	Multi-model access (Claude, GPT-5, Gemini) via GitHub subscription
Cline	Plan/Act mode, structured task completion
Cursor	Auto model selection, multi-model with Claude, GPT-5, Gemini
OpenClaw	Real-time WebSocket streaming, configurable thinking levels
OpenCode	Multi-backend agent (Anthropic, OpenAI, Google, Groq)
Qwen Code	Alibaba's AI coding agent, deep reasoning
Ollama	Local LLM inference, privacy-first, no subscription
LocalAI	Self-hosted AI models, full control

Switch providers with one click. No lock-in.

Why Mysti?

vs Copilot/Cursor	Mysti Advantage
Single AI	Multi-agent brainstorming — two AIs collaborate with 5 strategies
Locked to one provider	12 providers — Claude, Codex, Gemini, Copilot, Cline, Cursor, OpenClaw, Manus, OpenCode, Qwen, Ollama, LocalAI
Black box	Full permission control — read-only to full-access
Generic responses	16 personas — architect, debugger, security expert...
Manual workflow	Autonomous mode — AI works independently with safety controls
No cross-agent routing	@-mentions — route tasks to specific agents inline

---

See It In Action

Beautiful, modern chat interface with syntax highlighting, markdown support, and mermaid diagrams

Real-time task list rendering and progress tracking

---

Brainstorm Mode

Want a second opinion? Enable Brainstorm Mode and let two AI agents tackle your problem together. Choose any 2 of 12 agents from the settings panel.

5 Collaboration Strategies

Strategy	Roles	Best For
Quick	Direct synthesis	Simple tasks, fast answers
Debate	Critic vs Defender	Architecture decisions, trade-offs
Red-Team	Proposer vs Challenger	Security reviews, edge case discovery
Perspectives	Risk Analyst vs Innovator	Greenfield design, technology selection
Delphi	Facilitator vs Refiner	Complex problems, reaching consensus

Why Two AIs Beat One

Claude Code (Anthropic), Codex (OpenAI), Gemini (Google), GitHub Copilot, Cline, Cursor, OpenClaw, OpenCode, Qwen Code (Alibaba), Ollama, and LocalAI have different training, different strengths, and different blind spots. When any two work together:

• Each AI catches edge cases the other might miss
• Different perspectives lead to more robust solutions
• Together they debate, challenge each other, and synthesize the best solution

It's like having a senior dev and a tech lead review your code—except they actually discuss it first.

Convergence Detection

During discussions, Mysti tracks agent agreement and position stability. When auto-convergence is enabled, the discussion exits early once agents reach consensus — saving time without sacrificing quality.

Choose Your Team

Configure which two agents collaborate in the Settings Panel:

Combination	Best For
Claude + Codex	Deep analysis meets rapid iteration
Claude + Gemini	Thorough reasoning with fast validation
Claude + Copilot	Compare native Claude vs Copilot's multi-model approach
Cursor + Gemini	Multi-model flexibility with Google integration
OpenClaw + Claude	WebSocket streaming with deep reasoning
Qwen + Claude	Compare Alibaba and Anthropic reasoning
OpenCode + Gemini	Multi-backend flexibility with Google speed
Ollama + Claude	Local privacy meets cloud intelligence

Full Brainstorm documentation

Intelligent Plan Detection

When the AI presents multiple implementation approaches, Mysti automatically detects them and lets you choose your preferred path.

Requires at least 2 CLI tools installed. See Requirements.

---

Key Features

Autonomous Mode

Let the AI work independently with configurable safety controls:

• Safety Classifier: Three levels — safe (auto-approve), caution (mode-dependent), blocked (always deny)
• Three Safety Modes: Conservative, Balanced, Aggressive
• Learning Memory: Remembers your permission preferences and improves over time
• Continuation Modes: Goal-based or task-queue for extended autonomous sessions
• Audit Trail: Every autonomous decision is logged for review

Full Autonomous Mode documentation

@-Mention System

Route tasks to specific agents and reference files inline:

@claude Review this code for security issues
@src/auth.ts @gemini Suggest performance improvements for this file
@claude Write tests, then @codex optimize them

• File mentions: @filename adds transient context
• Agent mentions: @agent routes tasks to that provider
• Chaining: Later agents receive earlier agents' responses as context

Full @-Mention documentation

Context Compaction

Smart conversation management that prevents context overflow:

• Automatic: Triggers when token usage approaches the threshold (default 75%)
• Native support: Claude Code uses built-in /compact command
• Client-side: Other providers use intelligent message summarization
• Per-panel tracking: Each chat panel tracks usage independently

Full Compaction documentation

16 Developer Personas

Shape how your AI thinks. Select from specialized personas that change the AI's approach to your problems.

Persona	Focus
Architect	System design, scalability, clean structure
Debugger	Root cause analysis, bug fixing
Security-Minded	Vulnerabilities, threat modeling
Performance Tuner	Optimization, profiling, latency
Prototyper	Quick iteration, PoCs
Refactorer	Code quality, maintainability
+ 10 more...	Full-Stack, DevOps, Mentor, Designer...

Full Personas & Skills documentation

---

Quick Persona Selection

Select personas directly from the toolbar without opening panels.

---

Smart Auto-Suggestions

Mysti automatically suggests relevant personas and actions based on your message.

---

Conversation History

Never lose your work. All conversations are saved and easily accessible.

---

Quick Actions on Welcome

Get started fast with one-click actions for common tasks.

---

Extensive Settings

Fine-tune every aspect of Mysti including token budgets, access levels, and brainstorm mode.

---

Requirements

Already paying for Claude, ChatGPT, Gemini, or GitHub Copilot? You're ready to go.

Mysti works with your existing subscriptions—no additional costs!

CLI Tool	Subscription	Install
Claude Code (recommended)	Anthropic API or Claude Pro/Max	npm install -g @anthropic-ai/claude-code
GitHub Copilot CLI	GitHub Copilot Pro/Pro+/Business	npm install -g @github/copilot-cli
Gemini CLI	Google AI API or Gemini Advanced	npm install -g @google/gemini-cli
Codex CLI	OpenAI API	Follow OpenAI's installation guide
Cline	Depends on model provider	npm install -g cline
Cursor	Cursor subscription	curl https://cursor.com/install -fsS | bash
OpenClaw	OpenClaw account	npm install -g openclaw@latest && openclaw onboard --install-daemon
OpenCode	Provider API keys (Anthropic, OpenAI, etc.)	npm i -g opencode-ai@latest
Qwen Code	Qwen OAuth or API keys	npm install -g @qwen-code/qwen-code@latest
Ollama	Local (no subscription needed)	Install from ollama.com
LocalAI	Local (no subscription needed)	Install from localai.io

You only need one CLI to get started. Install any two to unlock Brainstorm Mode.

---

Quick Start

1. Install Mysti

Option A: Press Ctrl+P (Cmd+P on Mac), paste and run:

ext install DeepMyst.mysti

Option B: Install from VS Code Marketplace

2. Install a CLI Tool

# Claude Code (recommended)
npm install -g @anthropic-ai/claude-code
claude auth login

# Or GitHub Copilot CLI (access Claude, GPT-5, Gemini via GitHub)
npm install -g @github/copilot-cli
copilot  # then use /login command

# Or Gemini CLI
npm install -g @google/gemini-cli
gemini auth login

# Or Cursor
curl https://cursor.com/install -fsS | bash
agent login

# Or OpenClaw
npm install -g openclaw@latest && openclaw onboard --install-daemon
openclaw login

# Or OpenCode
npm i -g opencode-ai@latest
opencode auth login

# Or Qwen Code
npm install -g @qwen-code/qwen-code@latest
qwen  # then type /auth

For Brainstorm Mode, install any two CLI tools.

3. Open Mysti

• Click the Mysti icon in the Activity Bar, or
• Press Ctrl+Shift+M (Cmd+Shift+M on Mac)

4. Start Coding

Type your request and let the AI assist you!

---

Slash Commands

Access skills and actions quickly with the built-in slash command menu.

---

12 Toggleable Skills

Mix and match behavioral modifiers:

• Concise - Clear, brief communication
• Test-Driven - Tests alongside code
• Auto-Commit - Incremental commits
• First Principles - Fundamental reasoning
• Scope Discipline - Stay focused on the task
• And 7 more...

Full Personas & Skills documentation

---

Permission Controls

Stay in control of what the AI can do:

• Read-only - AI can only read, never modify
• Ask-permission - Approve each file change
• Full-access - Let the AI work autonomously

---

Configuration

Essential Settings

{
  "mysti.defaultProvider": "claude-code",
  "mysti.brainstorm.agents": ["claude-code", "google-gemini"],
  "mysti.brainstorm.strategy": "quick",
  "mysti.accessLevel": "ask-permission"
}

Provider Settings

Setting	Default	Description
mysti.defaultProvider	claude-code	Primary AI provider
mysti.claudePath	claude	Path to Claude CLI
mysti.codexPath	codex	Path to Codex CLI
mysti.geminiPath	gemini	Path to Gemini CLI
mysti.copilotPath	copilot	Path to Copilot CLI
mysti.clinePath	cline	Path to Cline CLI
mysti.cursorPath	agent	Path to Cursor CLI
mysti.openclawPath	openclaw	Path to OpenClaw CLI
mysti.opencodePath	opencode	Path to OpenCode CLI
mysti.qwenCodePath	qwen	Path to Qwen Code CLI
mysti.ollamaPath	ollama	Path to Ollama CLI
mysti.localaiPath	localai	Path to LocalAI CLI

Brainstorm Settings

Setting	Default	Description
mysti.brainstorm.agents	["claude-code", "openai-codex"]	Which 2 agents to use
mysti.brainstorm.strategy	quick	Strategy: quick, debate, red-team, perspectives, delphi
mysti.brainstorm.autoConverge	true	Auto-exit when agents converge
mysti.brainstorm.maxDiscussionRounds	3	Maximum discussion rounds

Autonomous Settings

Setting	Default	Description
mysti.autonomous.safetyMode	balanced	conservative, balanced, aggressive
mysti.autonomous.blockPatterns	[]	Custom patterns to always block

Compaction Settings

Setting	Default	Description
mysti.compaction.enabled	true	Enable context compaction
mysti.compaction.threshold	75	Compaction threshold (% of context window)

General Settings

Setting	Default	Description
mysti.accessLevel	ask-permission	File access level
mysti.agents.autoSuggest	true	Auto-suggest personas
mysti.agents.maxTokenBudget	0	Max tokens for agent context (0 = unlimited)

Full Provider documentation

---

Keyboard Shortcuts

Action	Windows/Linux	Mac
Open Mysti	Ctrl+Shift+M	Cmd+Shift+M
Open in New Tab	Ctrl+Shift+N	Cmd+Shift+N

---

Commands

Command	Description
Mysti: Open Chat	Open the chat sidebar
Mysti: New Conversation	Start fresh
Mysti: Add to Context	Add file/selection to context
Mysti: Clear Context	Clear all context
Mysti: Open in New Tab	Open chat as editor tab

---

Documentation

Guide	Description
Providers	All 12 providers — setup, models, features
Brainstorm Mode	5 strategies, convergence, team selection
Personas & Skills	16 personas, 12 skills, custom agents
Autonomous Mode	Safety system, memory, continuation modes
@-Mentions	Agent routing and file context
Compaction	Context management and summarization
Architecture	Technical internals and extension points
Features	Complete feature reference

---

Telemetry

Mysti collects anonymous usage data to improve the extension:

• Feature usage patterns
• Error rates
• Provider preferences

No code, file paths, or personal data is ever collected.

Respects VSCode's telemetry setting. Disable via:
Settings > Telemetry: Telemetry Level > off

---

Contributors

Thanks to everyone who has helped make Mysti better!

Want to join them? Check out the Contributing section below.

---

Star History

If Mysti has been useful to you, consider giving it a star — it helps others discover the project and keeps us motivated!

---

Contributing

We welcome contributions! Whether it's bug reports, feature requests, or code contributions.

• Good First Issues: Look for good first issue labels
• Development: Press F5 in VS Code to launch Extension Development Host
• Pull Requests: Fork, create a feature branch, and submit a PR

See CONTRIBUTING.md for detailed guidelines.

---

License

Apache License 2.0 — free to use, modify, and distribute, including for commercial purposes.
See the LICENSE file for full text.

---

Install • Report Issue • GitHub

Mysti — Built by DeepMyst Inc
_{Made with Mysti}