Web Interoperable Linked Knowledge
A pure command-line platform that treats AI agents as modular, composable "packages"—installable, versioned, and linkable—within a familiar, scriptable, and offline-capable CLI environment.
I'd like to thank libre chat, claude code, gemini cli and cursor for bringing this totally not AI slop into existence with me in the drivers seat.
Also the chances of someone else doing a better of this before me is pretty high. I'm poor and the token cost to build this with an actual good model (claude-code takes the w) honestly makes sweat. I doubt they were higher than me when I thought this out though.
Wilk fundamentally changes how humans interact with and leverage AI. Moving beyond proprietary UIs and cumbersome frameworks, Wilk provides a declarative operating system for building, managing, and sharing intelligent agents and complex AI workflows directly from the terminal.
Think of it as:
- Package manager for AI agents (like npm, but for intelligent assistants)
- Git for AI workflows (version, branch, and share agent configurations)
- tmux for AI conversations (persistent, resumable, multi-agent sessions)
The current AI landscape forces developers into an impossible choice:
❌ Heavy Frameworks (LangChain, AutoGen, CrewAI)
Powerful but require extensive Python boilerplate, complex infrastructure, and steep learning curves
❌ Closed-Source Chat Interfaces (ChatGPT, Claude, Gemini)
Accessible but proprietary black boxes with limited tooling, no composability, and online dependency
❌ Single-Purpose CLI Tools (Aider, Copilot CLI, Cursor)
Focused utility but lacking multi-agent orchestration, workflow composition, and ecosystem integration
✅ Wilk bridges this gap by providing enterprise-grade capabilities with terminal-native simplicity.
┌─────────────────────────────────────────────────────────────────────┐
│> What are the latest developments in quantum computing? │
└─────────────────────────────────────────────────────────────────────┘
🤔 Agent thinking... (3.2s · ↑ 1.8k tokens)
● Task(Quantum Computing Research)
│ ⎿ Web Search("quantum computing 2024 developments")
│ Found 15 results (enter to expand)
│ ⎿ Analyzing latest papers...
│ ✓ 8 papers analyzed
│ ⎿ Synthesizing findings...
│ ✓ Research complete
📊 Latest Quantum Computing Developments (2024)
[Detailed analysis with transparent reasoning...]
┌─────────────────────────────────────────────────────────────────────┐
│ >/agent add code-analyzer doc-generator │
└─────────────────────────────────────────────────────────────────────┘
✅ Added agents to current session: code-analyzer, doc-generator
┌─────────────────────────────────────────────────────────────────────┐
│> analyse this code base and generate docs │
└─────────────────────────────────────────────────────────────────────┘
🎭 Multi-Agent Orchestration Started
├─ Agent(code-analyzer) → Scanning ./src for patterns and issues
├─ Agent(doc-generator) → Creating documentation from analysis
│
● Task(Code Analysis)
│ ⎿ Scanning 23 TypeScript files...
│ ✓ Found 3 critical issues, 12 warnings
│ ⎿ Analyzing architecture patterns...
│ ✓ Identified 5 design patterns
│
● Task(Documentation Generation)
│ ⎿ Generating API documentation...
│ ✓ Created ./docs/API.md
│ ⎿ Creating architecture overview...
│ ✓ Created ./docs/ARCHITECTURE.md
│
✅ Multi-agent collaboration complete (1m 32s)
📊 Analysis: 3 critical issues found
📚 Documentation: 2 new files generated
- <100ms startup time (vs 2-5s for web LibreChat)
- SQLite storage (10-50x faster than MongoDB for CLI use)
- Local-first with offline capability
- Memory usage <50MB for typical sessions
# Install Wilk globally
npm install -g @wilk/cli
# Start interactive session
wilk# Just run wilk to start
$ wilk
┌──────────────────────────────────────┐
│ │
│ ██╗ ██╗██╗██╗ ██╗ ██╗ │
│ ██║ ██║██║██║ ██║ ██╔╝ │
│ ██║ █╗ ██║██║██║ █████╔╝ │
│ ██║███╗██║██║██║ ██╔═██╗ │
│ ╚███╔███╔╝██║███████╗██║ ██╗ │
│ ╚══╝╚══╝ ╚═╝╚══════╝╚═╝ ╚═╝ │
│ │
│ CLI-Native Agent Operating System │
│ │
└──────────────────────────────────────┘
Available agents: @research, @code-analyzer, @docs-writer
Memory loaded: 3 entries (847 tokens)
Type your message, or use @agent-name for specific agents
Press ↑/↓ for history, Tab for completion, Ctrl+C to exit
Tips for getting started:
1. Ask questions, edit files, or run commands
2. Be specific for the best results
3. Create WILK.md files to customize interactions
4. /help for more information
┌────────────────────────────────────────────────────────────────────────────┐
│ > Hello, help me understand this codebase │
└────────────────────────────────────────────────────────────────────────────┘
# Optional: Initialize project memory
┌────────────────────────────────────────────────────────────────────────────┐
│ >/init # Creates WILK.md for project context │
└────────────────────────────────────────────────────────────────────────────┘
# Create agents as needed
┌────────────────────────────────────────────────────────────────────────────┐
│ >/agent create my-assistant --provider openai --model gpt-4 │
└────────────────────────────────────────────────────────────────────────────┘
┌────────────────────────────────────────────────────────────────────────────┐
│ >/list-agents │
└────────────────────────────────────────────────────────────────────────────┘
Available agents: @my-assistant
# Multi-agent workflow to call additional agents into a session prompt
┌────────────────────────────────────────────────────────────────────────────┐
│ >/@code-analyzer @security-audit Review the authentication module │
└────────────────────────────────────────────────────────────────────────────┘
- Agent Management - Creating, configuring, and versioning agents
- Context Management - Memory, summarization, and context strategies
- Session Management - Persistent conversations with thinking preservation
- Command Reference - Complete command documentation (80+ commands)
- Agent Syntax - Agent interaction patterns and multi-agent coordination
- Memory System - Project/user/agent persistent memory
- Tool Integration - File operations, web search, code execution, RAG
- MCP Integration - Model Context Protocol server management
- Thinking Implementation - Agent reasoning transparency
- Task Execution Display - Hierarchical task visualization
- Node.js 18+
- SQLite 3.x
- TypeScript 5.x
# Clone the repository
git clone https://github.com/your-org/wilk.git
cd wilk
# Install dependencies
npm install
# Setup development environment
./scripts/agent-kiwi.sh dev-setup
# Run in development mode
npm run devWilk is designed for CLI-native performance and developer productivity:
┌─────────────────────────────────────────────────────────────┐
│ Wilk CLI Interface │
├─────────────────────────────────────────────────────────────┤
│ REPL │ Commands │ Agent Syntax │ Session Management │
├─────────────────────────────────────────────────────────────┤
│ Agent Orchestration Layer │
├─────────────────────────────────────────────────────────────┤
│ Agent Mgmt │ Context Mgmt │ Tool Integration │ Security │
├─────────────────────────────────────────────────────────────┤
│ Core Engine Layer │
├─────────────────────────────────────────────────────────────┤
│ LLM Clients │ Agent Runtime │ Tool Manager │ Memory System │
├─────────────────────────────────────────────────────────────┤
│ Storage Layer │
├─────────────────────────────────────────────────────────────┤
│ SQLite DB │ YAML Configs │ Agent Definitions │
└─────────────────────────────────────────────────────────────┘
Key Design Principles:
- CLI-first design with <100ms startup time
- Local-first with git-friendly configuration
- Modular agents as composable packages
- Transparent operations with full reasoning visibility
We welcome contributions! Wilk is built on the principle that AI agent development should be accessible, transparent, and community-driven.
- 🐛 Bug Reports - Help us improve stability and performance
- ✨ Feature Requests - Share ideas for new capabilities
- 🔧 Code Contributions - Submit PRs for bug fixes or features
- 📚 Documentation - Improve guides, tutorials, and examples
- 🤖 Agent Packages - Create and share reusable agents
- 🛠️ Tool Integration - Build new tool integrations
- Fork the repository
- Create a feature branch (
git checkout -b feature/amazing-feature) - Make your changes
- Add tests if applicable
- Commit your changes (
git commit -m 'Add amazing feature') - Push to the branch (
git push origin feature/amazing-feature) - Open a Pull Request
See CONTRIBUTING.md for detailed guidelines.
┌────────────────────────────────────────────────────────────────────────────┐
│ >/prompt add bug-checklist Check for null pointer exceptions, race │
│ conditions, and memory leaks. │
└────────────────────────────────────────────────────────────────────────────┘
✅ Added prompt: bug-checklist
┌────────────────────────────────────────────────────────────────────────────┐
│ >%bug-checklist │
└────────────────────────────────────────────────────────────────────────────┘
📝 Running bug-checklist on ./src/utils/
- Null pointer checks: OK
- Race conditions: 1 potential issue
- Memory leaks: None detected
┌────────────────────────────────────────────────────────────────────────────┐
│ >/prompt add code-review Review {{filename}} for {{review_type}} issues. │
│ Focus on {{language}} best practices. Report generated on {{current_date}} │
└────────────────────────────────────────────────────────────────────────────┘
✅ Added prompt: code-review
┌───────────────────────────────────────────────────────────────────────────────┐
│ >%code-review --filename auth.js --review_type security --language javascript │
└───────────────────────────────────────────────────────────────────────────────┘
📝 Running code-review on auth.js
🔍 Security Review for auth.js (JavaScript)
- Authentication flow: 2 potential vulnerabilities found
- Input validation: Missing validation on login endpoint
- Session management: Secure implementation detected
Report generated on Monday, January 15, 2024
┌────────────────────────────────────────────────────────────────────────────┐
│ >@security @performance @maintainability Review this pull request │
└────────────────────────────────────────────────────────────────────────────┘
● Task(Multi-Perspective Code Review)
├─ Agent(security) → Security analysis
├─ Agent(performance) → Performance bottlenecks
├─ Agent(maintainability) → Code quality assessment
│
● Task(Synthesis) → Combined recommendations ready
┌────────────────────────────────────────────────────────────────────────────┐
│ >Analyze the ./src directory for potential issues and generate updated │
│ API documentation. │
└────────────────────────────────────────────────────────────────────────────┘
🤖 code-analyzer: Scanning ./src for issues...
🔍 3 critical, 12 minor issues found
🤖 doc-generator: API documentation generated at ./docs/API.md
┌────────────────────────────────────────────────────────────────────────────┐
│ >#Default to using Python 3.11 for all new scripts │
└────────────────────────────────────────────────────────────────────────────┘
Where should this be saved?
1. Project memory (./WILK.md)
2. User memory (~/.wilk/memory/user.md)
Choice [1]: 1
✅ Added to project memory
┌────────────────────────────────────────────────────────────────────────────┐
│ >/memory │
└────────────────────────────────────────────────────────────────────────────┘
📁 Project Memory: ./WILK.md (1,247 tokens)
👤 User Memory: ~/.wilk/memory/user.md (567 tokens)
🤖 Agent Memory: 3 agents with learning data
┌────────────────────────────────────────────────────────────────────────────┐
│ >/search docker optimization │
└────────────────────────────────────────────────────────────────────────────┘
Found 2 agents:
- docker-optimizer (⭐ 4.6, 450 downloads)
- container-analyzer (⭐ 4.3, 320 downloads)
┌────────────────────────────────────────────────────────────────────────────┐
│ >/install docker-optimizer │
└────────────────────────────────────────────────────────────────────────────┘
✅ Installed agent: docker-optimizer
┌────────────────────────────────────────────────────────────────────────────┐
│ >/agent add docker-optimizer │
└────────────────────────────────────────────────────────────────────────────┘
✅ Added agents to current session: docker-optimizer
┌────────────────────────────────────────────────────────────────────────────┐
│ >Analyze and optimize our Dockerfile │
└────────────────────────────────────────────────────────────────────────────┘
🔍 Analyzing Dockerfile...
✅ Found 3 optimization opportunities
📊 Reduced image size by 45%
🚀 Published optimization report to team registry
OR
┌────────────────────────────────────────────────────────────────────────────┐
│ >@docker-optimizer Analyze and optimize our Dockerfile │
└────────────────────────────────────────────────────────────────────────────┘
🔍 Analyzing Dockerfile...
✅ Found 3 optimization opportunities
📊 Reduced image size by 45%
🚀 Published optimization report to team registry
┌────────────────────────────────────────────────────────────────────────────┐
│ >/prompt upload docker-best-practices --description "Docker optimization │
│ checklist for production deployments" │
└────────────────────────────────────────────────────────────────────────────┘
📤 Uploaded prompt to community repository
┌────────────────────────────────────────────────────────────────────────────┐
│ >/stats --global --category "devops" │
└────────────────────────────────────────────────────────────────────────────┘
Global DevOps Agent Statistics:
- Total agents: 47
- Average rating: 4.3/5
- Most popular: ci-cd-optimizer (2,100 downloads)
- Trending: docker-optimizer (+15% this week)
┌────────────────────────────────────────────────────────────────────────────┐
│ >/mcp list │
└────────────────────────────────────────────────────────────────────────────┘
Configured MCP servers:
- github-server (running)
- filesystem-server (stopped)
┌────────────────────────────────────────────────────────────────────────────┐
│ >/mcp start filesystem-server │
└────────────────────────────────────────────────────────────────────────────┘
🟢 Started filesystem-server
┌────────────────────────────────────────────────────────────────────────────┐
│ >/permissions set code-analyzer filesystem read ./src │
└────────────────────────────────────────────────────────────────────────────┘
✅ Set permission: code-analyzer can read ./src
┌────────────────────────────────────────────────────────────────────────────┐
│ >/status │
└────────────────────────────────────────────────────────────────────────────┘
Wilk v2.3.0
Active agents: code-analyzer, doc-generator, test-coverage-reporter
API: Connected (OpenAI, Ollama)
Session: Active (42 min)
Tokens: 20k used
Memory: 320MB used
┌────────────────────────────────────────────────────────────────────────────┐
│ >/doctor │
└────────────────────────────────────────────────────────────────────────────┘
🔍 Wilk Health Check
✅ Installation: OK
✅ LLM Connection: OK
⚠️ Warning: test-coverage-reporter update available
┌────────────────────────────────────────────────────────────────────────────┐
│ >/api-config install community/google-cloud-apis │
└────────────────────────────────────────────────────────────────────────────┘
✅ Installed API configuration: google-cloud-apis
┌────────────────────────────────────────────────────────────────────────────┐
│ >/permissions set security-audit filesystem read ./src │
└────────────────────────────────────────────────────────────────────────────┘
✅ Set permission: security-audit can read ./src
┌────────────────────────────────────────────────────────────────────────────┐
│ >/permissions set security-audit network allowed_hosts api.github.com │
└────────────────────────────────────────────────────────────────────────────┘
✅ Set permission: security-audit can access api.github.com
┌────────────────────────────────────────────────────────────────────────────┐
│ >/logs security-audit │
└────────────────────────────────────────────────────────────────────────────┘
Recent logs for security-audit:
[2024-01-15 14:30:12] [INFO] Agent started
[2024-01-15 14:30:15] [DEBUG] Scanning ./src for vulnerabilities
[2024-01-15 14:30:18] [WARN] Found potential SQL injection in auth.js:45
┌────────────────────────────────────────────────────────────────────────────┐
│ >/search code review │
└────────────────────────────────────────────────────────────────────────────┘
Found 3 agents:
- code-reviewer (⭐ 4.8, 1200 downloads)
- secure-review-bot (⭐ 4.7, 900 downloads)
- pr-helper (⭐ 4.5, 800 downloads)
┌────────────────────────────────────────────────────────────────────────────┐
│ >/search --category devops --rating >4.0 │
└────────────────────────────────────────────────────────────────────────────┘
Found 2 agents in category 'devops' with rating >4.0:
- ci-cd-optimizer (⭐ 4.9)
- docker-linter (⭐ 4.2)
┌────────────────────────────────────────────────────────────────────────────┐
│ >/browse --category web-development │
└────────────────────────────────────────────────────────────────────────────┘
Browsing web-development agents:
- react-helper
- api-doc-gen
- frontend-tester
┌────────────────────────────────────────────────────────────────────────────┐
│ >/browse --trending --last-week │
└────────────────────────────────────────────────────────────────────────────┘
Trending agents this week:
- code-reviewer
- ci-cd-optimizer
- markdown-docs
┌────────────────────────────────────────────────────────────────────────────┐
│ >/showcase --featured │
└────────────────────────────────────────────────────────────────────────────┘
Featured agents:
1. code-reviewer ⭐ 4.9 — Automated code review for multiple languages
2. doc-generator ⭐ 4.8 — Generate project documentation from code
3. ci-cd-optimizer ⭐ 4.7 — Optimize your CI/CD pipelines
┌────────────────────────────────────────────────────────────────────────────┐
│ >/publish my-agent --description "TypeScript code analyzer" │
└────────────────────────────────────────────────────────────────────────────┘
🚀 Published agent 'my-agent' to the community registry
┌────────────────────────────────────────────────────────────────────────────┐
│ >/publish --private --team my-company/internal-tools │
└────────────────────────────────────────────────────────────────────────────┘
🔒 Published agent privately to team: my-company/internal-tools
┌────────────────────────────────────────────────────────────────────────────┐
│ >/stats code-reviewer │
└────────────────────────────────────────────────────────────────────────────┘
Statistics for code-reviewer:
- Downloads: 1,247
- Rating: 4.8/5 (89 reviews)
- Last updated: 2 days ago
- Compatibility: 98%
┌────────────────────────────────────────────────────────────────────────────┐
│ >/api-key add openai sk-1234567890abcdef │
└────────────────────────────────────────────────────────────────────────────┘
✅ Added OpenAI API key
┌────────────────────────────────────────────────────────────────────────────┐
│ >/api-key add github ghp_abcdef1234567890 │
└────────────────────────────────────────────────────────────────────────────┘
✅ Added GitHub API key
┌────────────────────────────────────────────────────────────────────────────┐
│ >/api-key list │
└────────────────────────────────────────────────────────────────────────────┘
Configured API keys:
- openai: sk-****...def
- github: ghp_****...890
┌────────────────────────────────────────────────────────────────────────────┐
│ >/api-key remove github │
└────────────────────────────────────────────────────────────────────────────┘
🗑️ Removed GitHub API key
┌────────────────────────────────────────────────────────────────────────────┐
│ >/debug code-analyzer --debug --trace-level verbose │
└────────────────────────────────────────────────────────────────────────────┘
🔍 Debug mode enabled for agent 'code-analyzer'
Trace level: verbose
All actions, tool calls, and LLM messages will be logged in real time.
┌────────────────────────────────────────────────────────────────────────────┐
│ >/logs code-analyzer │
└────────────────────────────────────────────────────────────────────────────┘
Recent logs for agent 'code-analyzer':
[2024-01-15 15:20:12] [INFO] Agent started
[2024-01-15 15:20:13] [DEBUG] Loaded tool: file-analyzer
[2024-01-15 15:20:14] [TRACE] LLM input: "Analyze ./src for issues"
[2024-01-15 15:20:15] [TRACE] LLM output: "Found 3 critical issues"
┌────────────────────────────────────────────────────────────────────────────┐
│ >/trace code-analyzer │
└────────────────────────────────────────────────────────────────────────────┘
🔬 Starting execution trace for agent 'code-analyzer'...
Step 1: Received input "Analyze all TypeScript files"
Step 2: Invoked tool 'file-analyzer'
Step 3: LLM response: "Analysis complete"
Step 4: Finished execution
Trace complete. Use /logs code-analyzer for full details.
┌────────────────────────────────────────────────────────────────────────────┐
│ >/diagnose │
└────────────────────────────────────────────────────────────────────────────┘
🩺 Running diagnostics...
✅ Core system: OK
✅ LLM connectivity: OK
✅ File system access: OK
✅ Agent registry: OK
No issues detected. Wilk is ready to use!
Wilk is designed with security and privacy as core principles:
- 🔐 Local-first architecture - Your data stays on your machine
- 🛡️ Permission system - Fine-grained control over agent capabilities
- 📋 Audit logging - Complete transparency into agent actions
- 🔍 Sandboxed execution - Safe tool execution environment
- 🚫 No telemetry - Zero data collection by default
- CLI interface with interactive REPL
- Agent management and versioning
- Memory system integration
- Basic tool integration
- Session persistence
- Agent thinking transparency
- Task execution display
- Agent marketplace and registry
- Community agent packages
- Agent templates and scaffolding
- Advanced multi-agent workflows
- Team collaboration
- Enterprise security controls
- Compliance automation (SOC2, GDPR)
- Advanced analytics and monitoring
- IDE extensions (VS Code, Vim, Emacs)
- CI/CD integrations
- Cloud deployment options
- Enterprise on-premises support
Wilk builds on the excellent work of:
- LibreChat - Our foundational architecture for agent management and conversation handling
- LangChain - Inspiration for tool integration patterns
- Model Context Protocol (MCP) - Standard for tool integration
- The open-source AI community - For pushing the boundaries of what's possible
This project is licensed under the MIT License - see the LICENSE file for details.
- 📧 Email: hello@wilk.dev
- 💬 Discord: Join our community
- 🐛 Issues: GitHub Issues
- 💡 Discussions: GitHub Discussions
"CLI-native AI that doesn't suck."
⭐ Star this repo if Wilk helps you save time with AI!