OpenClacky is building the cost-efficient execution layer for agentic work. Agents are moving from short demos into persistent workers that read files, use tools, control browsers, remember context, and execute workflows. As usage grows, the bottleneck shifts from raw model capability to cost-per-completed-task.
OpenClacky starts with a token-efficient, open-source agent runtime, then expands into AI Keys, reusable Skills, and enterprise workflow deployment. The product wedge includes a local AI workbench, Web UI, CLI, memory, Skills, scheduled tasks, IM integration, browser automation, cache-aware execution, and model routing.
Public signals: MIT open source, 946 GitHub stars, BYOK/model freedom, Skill self-evolution, 16 core tools, and existing backing listed publicly from MiraclePlus, ZhenFund, Yingdong Capital, Variable Capital, Sequoia China, and Hillhouse Ventures.
Public materials: https://www.openclacky.com and https://github.com/clacky-ai/openclacky