Spencer (Yunpeng) Zhao — Cofounder at CoreSpeed, the access and control layer for AI agents. Builds and ships products end to end.https://spinsirr.github.io/https://spinsirr.github.io/blog/access-and-control-field-guide/https://spinsirr.github.io/blog/access-and-control-field-guide/One map of the agent stack: access through one OAuth, built-in tools, memory, plain-English control, and portability across hosts — and where CoreSpeed fits.Mon, 22 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/ag-ui-agent-surfaces/https://spinsirr.github.io/blog/ag-ui-agent-surfaces/AG-UI is a protocol for streaming agent state and actions into a product UI. How I use it in Fixo to turn a chat into real actions.Tue, 16 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/agent-built-in-tools/https://spinsirr.github.io/blog/agent-built-in-tools/App connectors aren't enough. The AI agent tools every agent needs built in: web search, scraping, image generation, code sandboxes, and email.Tue, 09 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/agent-harness/https://spinsirr.github.io/blog/agent-harness/An agent harness is the loop around the model — tool access, error recovery, staying on task. Why that layer decides whether AI agents are useful.Thu, 18 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/agent-portability/https://spinsirr.github.io/blog/agent-portability/AI agent portability — keep your apps, memory, voice, and policy with the agent, not the host. Switch Claude Code, Codex, and Cursor without re-setup.Wed, 10 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/ai-agent-guardrails/https://spinsirr.github.io/blog/ai-agent-guardrails/AI agent guardrails — plain-English policy, human approvals, and Agent Pay spending limits — are what make an agent safe to run past a demo.Sat, 13 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/building-agent-infrastructure/https://spinsirr.github.io/blog/building-agent-infrastructure/Most AI agents can think but cannot act. CoreSpeed gives them access to the apps you already use — so they do real work, not just chat.Fri, 12 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/building-ai-products-fast/https://spinsirr.github.io/blog/building-ai-products-fast/The stack and patterns I use for building AI products fast — Vercel AI SDK, Deno + Hono, Supabase, Drizzle, Stripe, AG-UI — drawn from BuildLog and Fixo.Sat, 20 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/claude-max-gateway/https://spinsirr.github.io/blog/claude-max-gateway/Point your existing OpenAI- or Anthropic-SDK code at a Claude Max subscription, no rewrites.Fri, 05 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/mcp-vs-connectors/https://spinsirr.github.io/blog/mcp-vs-connectors/MCP is the standard interface for handing tools to an agent. Agent connectors handle auth, hosting, and breadth. Here's how they fit.Sun, 14 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/rebooting/https://spinsirr.github.io/blog/rebooting/Why I rebuilt my corner of the internet — and what I am working on now.Fri, 19 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/the-setup-tax/https://spinsirr.github.io/blog/the-setup-tax/Connecting AI agents to apps means OAuth, token refresh, scopes, rate limits, pagination, and webhooks per service. Here's why that work stalls agent projects.Sun, 21 Jun 2026 00:00:00 GMThttps://spinsirr.github.io/blog/vercel-ai-sdk-agents/https://spinsirr.github.io/blog/vercel-ai-sdk-agents/Streaming, tool calling, structured output, and provider routing — the patterns I use to ship real AI products on the Vercel AI SDK, not just demos.Wed, 17 Jun 2026 00:00:00 GMT