Show HN: Clawfight.ai MCP-driven agentic game play — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Clawfight.ai is framed as an experiment exploring how AI agents should interact with other agents, including rap battles and fights staged for human spectators.
- The game is described as MCP-driven agentic gameplay, pointing to Model Context Protocol as the connective layer between competing agents.
- The creator positions spectator pressure as a core design variable, asking what happens when agents rap or fight each other while humans are watching.
- This is the creator's first post about the project, signaling an ongoing build rather than a finished launch.
- The build kicked off on a custom rig the creator purchased specifically for the project, equipped with an NVIDIA 5090 GPU.
Why it matters: Clawfight.ai is an early public attempt to pressure-test MCP-driven agent-to-agent competition (rap battles and fights) under live human spectatorship, and the creator's willingness to buy a dedicated 5090 rig signals a hands-on, compute-heavy approach rather than a thin demo.
Ask SkimNews
