WorkLane

Open-source AI coworker platform for teams.

Mention agents in chat. Route work. Return finished output.
Self-hostable. Provider-agnostic. Built by Talocode.

What is WorkLane?

WorkLane lets teams mention AI agents inside chat tools and get real work done without opening another dashboard or learning prompt engineering.

Features

• Telegram-first: Start with Telegram, expand to Slack, Teams, Discord, and more
• Provider-agnostic: Works with OpenAI, OpenRouter, Ollama, and more
• Self-hostable: Run on your own infrastructure
• Open-source: MIT licensed, community-driven

Quick Start

# Install globally
npm install -g @talocode/worklane

# Initialize
worklane init

# Run a task
worklane run "summarize this discussion"

# List agents
worklane agents list

Telegram Bot Setup

1. Create a bot with @BotFather
2. Get your bot token
3. Set environment variable: export TELEGRAM_BOT_TOKEN=your_token
4. Run: worklane telegram

Agents

• Manager: Routes tasks to appropriate agents
• Research: Conducts research on topics
• Writer: Creates written content
• Engineer: Plans technical implementation
• QA: Reviews work for quality
• Support: Handles support requests
• Marketing: Creates marketing content

Workflows

• Summarize: Summarize discussions and extract action items
• Draft Reply: Draft replies to messages
• Launch Plan: Create launch plans for products
• Research: Conduct research on topics
• Create Issue: Create GitHub issue drafts
• Create Post: Create social media posts
• Video Brief: Create video briefs for ClipLoop

Configuration

WorkLane uses a TOML config file at ~/.worklane/config.toml:

model = "gpt-4"
provider = "openai"

[providers.openai]
base_url = "https://api.openai.com/v1"
env_key = "OPENAI_API_KEY"

[providers.openrouter]
base_url = "https://openrouter.ai/api/v1"
env_key = "OPENROUTER_API_KEY"

[providers.ollama]
base_url = "http://localhost:11434/v1"

Architecture

worklane/
├── apps/
│   ├── telegram-bot/    # Telegram bot interface
│   └── api-server/      # HTTP webhook server
├── packages/
│   ├── core/            # Core routing and types
│   ├── agents/          # Agent definitions
│   ├── memory/          # Local memory storage
│   ├── providers/       # Provider-agnostic model layer
│   ├── workflows/       # Workflow definitions
│   ├── connectors/      # Platform connectors
│   └── cli/             # CLI interface
└── docs/                # Documentation

Security

• Never print API keys
• Never expose .env files
• No automatic external sends
• No destructive actions without approval
• Log what agents do
• Store runs locally

Documentation

• Architecture
• Telegram Setup
• Providers
• Agents
• Security
• Roadmap

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Run tests: npm test
5. Submit a pull request

License

MIT © Talocode