An enterprise platform for building and managing "digital employees," StaffDeck transforms professional experience and business processes into reusable, evolvable organizational assets. It targets enterprises and institutions seeking to elevate AI from personal productivity to organizational capability, enabling continuous operation, task automation, and expertise preservation.

How It Works

StaffDeck employs state-machine-driven procedural skills, generated from natural language, for accurate complex process execution. It features document-structure-aware knowledge retrieval, enabling digital employees to navigate and locate information step-by-step across documents. Autonomous execution via HTTP APIs, MCP, and scheduled tasks closes the improvement loop with long-term memory, complete traces, and feedback analysis.

Quick Start & Requirements

• Primary install/run command: Clone repository, set up Python 3.11+ and Node.js 20+, install dependencies (pip install -e "backend[dev]", npm --prefix frontend-enterprise ci), configure backend/.env with model endpoint/API key, and launch using scripts/dev_up.sh --detach (or PowerShell equivalent).
• Prerequisites: Python 3.11+, Node.js 20+, npm, an OpenAI-compatible Chat Completions endpoint and API key. No CUDA requirement for the application itself; hardware depends on the model service.
• Links: StaffDeck Official Website (implied by desktop downloads), GitHub Repository.

Highlighted Details

• Offers desktop downloads for macOS, Windows, and Linux.
• Features an "Agent-Friendly Quick Deploy" for integration with AI coding assistants like Cursor, Claude Code, or Codex.
• Core workflows include defining digital employees, configuring capabilities (knowledge bases, SOPs, tools), starting sessions, executing tasks with observable events, and enabling continuous improvement via memory and feedback.

Maintenance & Community

Community channels include WeChat, Feishu, and Discord. Contributions are welcomed via GitHub Issues for bugs/feature requests and focused pull requests for code and documentation improvements.

Licensing & Compatibility

Licensed under the GNU Affero General Public License v3.0 (AGPL-3.0). This strong copyleft license requires that any modifications or derivative works distributed over a network must also be made available under the AGPL-3.0. Compatibility for commercial use or linking with closed-source projects may be restricted due to these terms.

Limitations & Caveats

Model responses can be incorrect or inconsistent, and knowledge retrieval quality depends heavily on source data and model capabilities. External tools and generated runners may have real-world side effects. Scheduled tasks require a continuously running worker. This platform is not a substitute for professional review in regulated fields and requires appropriate authorization, privacy protection, and human oversight for sensitive data or decisions.