Plannix Build
Plannix Build is a powerful and extensible coding agent. Use it via an interactive TUI, headlessly in scripts or bots, or through the Agent Client Protocol (ACP) in other apps.
The TUI provides a rich, mouse-interactive, fullscreen experience for coding with agents.
Install
Start an interactive session
Navigate to your repository and launch Plannix. The TUI detects your project configuration, language tools, and git state automatically:
Run headlessly
Execute one-off coding prompts, automated bug fixes, or code reviews without opening the interactive UI:
Custom models
Plannix Build defaults to Plannix 3.6 Ultra. You can switch to other models anytime with the --model parameter or directly inside the interactive session:
- plannix-3.6-ultra: Frontier coding and multi-agent reasoning. 1,000,000 token context window.
- plannix-3-ultra: General text synthesis, RFC authoring, and code review.
- plannix-2.5-instant: Ultra-fast live autocomplete served on Groq LPU™ clusters.
Use Plannix 3.6 on the API
Build agentic pipelines with the official Plannix SDK or any OpenAI-compatible client library:
Features
Explore powerful agentic capabilities built into Plannix Build:
Extend Plannix with custom domain knowledge and npm/marketplace bundles.
Define persistent repository rules, coding styles, and project conventions.
Connect external databases, developer tools, and APIs via Model Context Protocol.
Interactive architectural planning and approval before agent modifies code.
Spawn parallel worker agents for codebase surveys and deep debugging.
Safely isolate agent edits in dedicated git worktrees without touching working branch.
Next
Ready to dive deeper into Plannix Build? Check out these essential guides:
- Modes and Commands — Learn how to switch between Normal, Plan, and Auto modes.
- Keyboard Shortcuts — Speed up navigation and session controls inside the TUI.
- Settings & Configuration — Customize environment variables, API endpoints, and telemetry.