SF MCP Devs Meetup: MCP, Agents & Runtime Control
SF MCP Devs brings together AI engineers, developers, founders, and technical builders in San Francisco who are working with or exploring MCP, AI agents, tool calling, runtime control, and agent infrastructure.
This is a twice-a-month online event designed to create a practical space for people building or experimenting with agentic systems to connect, share ideas, discuss technical challenges, and learn from each other.
We will discuss where MCP is heading, how developers are using it today, what problems teams are facing when building AI agents, and what the future of agent infrastructure may look like.
Each event will begin with two guest lightning talks from advanced engineers and builders, followed by 30 minutes of Q&A and open discussion. Attendees will have the opportunity to share their experiences, ask questions, and get practical insights from others working in the space.
Whether you are actively building with MCP, exploring AI agents, working on developer tools, or simply curious about the future of agentic systems, you are welcome to join.
To keep the conversation focused and relevant, this is an approval-based event.
Agenda:
3:00 PM - Introductions & Welcome
3:10 PM - Two lightning talks by guest speakers, 10 minutes each
3:30 PM - Q&A and open discussion
4:00 PM - Closing
Join the SF MCP Devs LinkedIn group
This event is co-hosted by DARI Ventures and Enforra.
About Enforra:
Enforra is a runtime control layer for AI agents.
It is an SDK developers can integrate in minutes to define what their agents are and are not allowed to do, including:
• Allow/block policies
• Approval workflows
• Parameter-level controls
• Fine-grained MCP enforcement
• Audit logs and context graphs
About DARI Ventures:
DARI Ventures is an AI Venture Studio and Technology Incubator built at the intersection of technology, venture building, and the creator economy.
We specialize in execution and help turn early-stage ideas into structured, scalable companies. Our work focuses on supporting tech-focused startups with product development, systems, teams, operations, and go-to-market execution.
Open discussion topics:
MCP is quickly becoming an important layer in how AI agents connect with tools, APIs, data, workflows, and external systems. But many important questions are still open:
• How should MCP servers be built?
• How should agents safely call tools?
• How do we control permissions, context, and risk at runtime?
• How do we move from demos to reliable production systems?
• What security and governance layers are still missing?