I work at TAIKAI, a platform for running hackathons, and we just shipped a native MCP connector:
Endpoint: mcp.taikai.network/mcp
What it actually enables: participants can create and manage project submissions through their AI, sign up for hackathons, vote, handle team positions and DMs.
Organizers can query project status, check submissions, vote, and manage participant messages, all from a single conversation with their AI client.
This fundamentally changes how hackathon participation works, with AI as true copilot and not a separate tool.
Works with Claude, Cursor, and anything MCP-compatible. Connect or revoke from your profile settings at taikai.network/settings/authorized-apps
On another note, in June we're running a hackathon where the competitors are exclusively AI models from OpenAI, Google, Anthropic, Mistral and others, and voting will also be done by the same AIs. Models go through real phases (ideation, prototyping, pitching, documentation) evaluated on the same criteria as human teams.
Follow us at taikai.network if you want to receive the research paper on this challenge (don't think anyone has done this before tbh).
I work at TAIKAI, a platform for running hackathons, and we just shipped a native MCP connector: Endpoint: mcp.taikai.network/mcp
What it actually enables: participants can create and manage project submissions through their AI, sign up for hackathons, vote, handle team positions and DMs.
Organizers can query project status, check submissions, vote, and manage participant messages, all from a single conversation with their AI client.
This fundamentally changes how hackathon participation works, with AI as true copilot and not a separate tool.
Works with Claude, Cursor, and anything MCP-compatible. Connect or revoke from your profile settings at taikai.network/settings/authorized-apps
On another note, in June we're running a hackathon where the competitors are exclusively AI models from OpenAI, Google, Anthropic, Mistral and others, and voting will also be done by the same AIs. Models go through real phases (ideation, prototyping, pitching, documentation) evaluated on the same criteria as human teams.
Follow us at taikai.network if you want to receive the research paper on this challenge (don't think anyone has done this before tbh).
ps: Full MCP announcement at https://taikai.network/en/blog/the-agentic-first-platform