Frame is a platform designed to bring structure, context, and organization to AI-assisted software development, particularly for larger projects involving multiple AI agents like Claude Code, Codex CLI, and Gemini CLI. It addresses common issues such as context loss, lack of project memory, inconsistent project structures, and session bleed-over. Frame benefits developers by streamlining onboarding, improving collaboration, and providing a unified environment for managing AI-driven workflows.

How It Works

Frame establishes a standardized project structure with key files: AGENTS.md for project rules, STRUCTURE.json for module mapping and fast file lookup via an intentIndex, PROJECT_NOTES.md for persistent architectural decisions, and tasks.json for task tracking. Each project operates within an isolated, context-specific session, ensuring no cross-contamination between projects. This standardized approach allows any AI tool to immediately understand the project's layout and state. The intentIndex significantly speeds up AI agent operations by mapping concepts directly to relevant files, eliminating broad codebase scans.

Quick Start & Requirements

• Installation: Clone the repository (git clone https://github.com/kaanozhan/Frame.git), navigate into the directory (cd Frame), install dependencies (npm install), and run the development server (npm run dev). Pre-built binaries are available on the releases page for macOS, Windows, and Linux.
• Prerequisites: Node.js 16+ and npm are required. Users must also have at least one AI CLI tool installed: Claude Code, Codex CLI, or Gemini CLI.
• Links: GitHub Repository

Highlighted Details

• Multi-Terminal Interface: Supports up to 9 terminals in a single window with resizable grid layouts and real PTY support via node-pty.
• Integrated Project Management: Features a task panel for visual tracking, a GitHub panel for issues and PRs, and a Git panel for branch management.
• Multi-AI Support: Seamlessly switch between Claude Code, Codex CLI, and Gemini CLI, with automatic context injection tailored to the active project.
• Fast File Lookup: Utilizes STRUCTURE.json's intentIndex for direct concept-to-file mapping, optimizing AI agent performance.

Maintenance & Community

The project includes a roadmap detailing "Done," "In Progress," and "Planned" features, indicating active development. Contributing guidelines are provided, encouraging community involvement. No specific community channels (e.g., Discord, Slack) or notable sponsorships are mentioned in the README.

Licensing & Compatibility

Frame is released under the MIT License, which generally permits broad use, modification, and distribution, including for commercial purposes, with minimal restrictions.

Limitations & Caveats

Current limitations stem from features listed as "In Progress" on the roadmap. Prompt history is not yet fully developed into a persistent developer style profile, and a web-based platform (Frame Server) is not yet available. Remote development capabilities via SSH are also planned but not yet implemented.