OpenFang

OpenFang runs recurring AI tasks on your own machine and reports their results to a desktop dashboard. It's for people who want research, monitoring or business automation to continue on a schedule without starting each job in chat.

• Scheduled task packages: Built-in agents called Hands produce cited research, monitor targets and build knowledge graphs, find and score prospects, or track forecasts against outcomes.
• Content and browser work: Agents can turn long videos into captioned shorts, manage an X account and complete browser workflows; the browser agent requires approval before purchases.
• Memory across channels: Agents retain conversation context and connect to Telegram, Discord, Slack and WhatsApp, so interactions can continue across messaging services.
• Custom automation: Create your own Hands, combine agents in workflows with conditional steps and loops, and connect external tools through MCP.
• Execution controls: Tool execution uses a WebAssembly sandbox, file access stays within the workspace, and prompt injection scanning and a linked audit trail support oversight of agent actions.

OpenFang is open source under Apache 2.0 and runs on macOS, Windows and Linux as a single binary with a native desktop app. Ollama, vLLM and LM Studio support local inference; Anthropic, Gemini and OpenAI are cloud alternatives. The runtime and memory run locally, while cloud model requests go to the selected provider. Web research, online publishing and messaging integrations use external services. The project warns of rough edges and breaking changes between releases.