GitHub Copilot is an AI assistant for software development. It can suggest code, explain repositories, edit files, review changes, and work on assigned tasks that developers can inspect as pull requests.
← Back to Category
AI Coding
Claude Deck
A self-hosted workspace for observing coding-agent sessions, editing local agent configuration and reviewing GitHub work, with mixed-harness teams, messaging and operator controls.
Directory information is for reference. Check the official site for current features, pricing and availability. Listing does not imply hands-on testing or a purchase recommendation.
About this product
Claude Deck is intended for operators who already have repositories, local agent CLIs and model credentials. It brings work tracking, terminals, team communication and configuration into one workspace while GitHub remains the record of issues and pull requests. Manual sessions are available without enabling automatic dispatch.
Key Features
- Inspect and edit Claude Code JSON or Codex TOML settings, including profiles and runtime options.
- Discover agent sessions running in tmux with Agent Bridge and view or interact with multiple terminal panes.
- Use Agent Mail and reusable team rosters for context requests, handoffs and local coordination.
- Review MCP servers, skills, plugins, transcripts and usage information; coverage varies by harness.
Use cases
- Inspect ownership, waiting reasons, pull requests and attempts requiring attention across a GitHub backlog.
- Review sessions, settings and team messages from several agent harnesses in a single workspace.
Who it is for
- For technical operators managing existing repositories and multiple local coding agents with human review.
Requirements
- The installation guide requires Python 3.11+, Node.js 22.12+ and at least one supported agent CLI on the same host.
- Bring your own model credentials and repository environment; session discovery operates on the relevant tmux sessions.
Quick Start
- Prepare Python, Node.js and a supported agent CLI in the same environment as the agent credentials.
- Clone the repository and follow its Installation script, backing up your configuration before the first changes.
- Start with Overview or manual sessions, verify repository and harness status, then configure teams, messaging and automation policies as needed.
Limitations and considerations
- UI changes can modify real local agent configuration files, so inspect changes and retain backups.
- The installation guide does not support Docker; some team pane-binding and permission capabilities require Linux, and harness coverage varies.
License and links
Code license: MIT
Sources
Source review: 2026-11-07