Introduction
Welcome to Knify - swiss knife for your agentic development.
What is Knify?
Knify lets you take the AI agents you built locally (in Cursor, Claude Code, OpenAI Codex, or Gemini) and run them in production. Build your agent once, then deliver it to your users anywhere.
Build → Upload → Run → Scale
Core Concepts
Jobs
Execute agentic workloads in isolated sandboxes. Each job runs your agent with full persistence and state management.
Workspaces
Upload your project files and configurations. Your agent runs with the same context it had locally.
Templates
Pre-configured sandbox environments with all dependencies installed. Optimized for fast startup and reliable execution.
Getting Started
Obtain your API key to authenticate requests.
Quickstart GuideGet up and running in 5 minutes.
Run your first agentic workload via the API.
Working with JobsLearn how to create and manage jobs.
Deploy your agent's workspace for production use.
Managing WorkspacesUpload and manage your project files.
Why Knify?
- Build Locally, Run Anywhere: Use your favorite AI coding tool, then deploy with one API call
- Persistent State: Conversations, files, and memory persist across job executions
- Fast Sandboxes: Isolated execution environments with ~150ms startup time
- Continuation Support: Resume conversations and build multi-turn workflows
- No Lock-In: Bring your own agent, built however you like
Explore the Docs
API Reference
Complete reference for all endpoints, parameters, and responses.
View API ReferenceLink to a page in the guide
Authentication
Learn how to secure your API requests with API keys.
View Authentication GuideLink to a page in the guide
Use Cases
Real-world examples and templates for common workflows.
View Use CasesLink to a page in the guide
Advanced Usage
Multi-turn workflows, webhooks, and orchestration patterns.
View Advanced GuideLink to a page in the guide
Templates
Optimize performance with pre-configured sandbox templates.
View Templates GuideLink to a page in the guide