Pick Your Stack
5 questions → recommended agent + MCP + framework stack tailored to your situation.
Calculators, comparators, wizards. Pick your stack, compare costs, decide protocols. Interactive — no scrolling through articles.
5 questions → recommended agent + MCP + framework stack tailored to your situation.
Tokens × calls × days → projected monthly cost. Compare Claude vs GPT vs Gemini.
Side-by-side: latency, context window, pricing, capabilities, by task type.
Pick 2–3 MCPs → see existing recipes that combine them.
Should you use MCP, Skills, Apps SDK, or roll your own? Interactive flowchart.
Paste a GitHub repo URL → get embeddable SVG badge. Public good.
Articles tell you. Calculators answer. Wizards decide. Comparators show.
Every Lab tool exists to answer a real question that comes up over and over in agentic AI work. They run client-side (no server cost, no data leaves your browser), load instantly, and never require a login.
Phase 1 ships Pick Your Stack. Phase 2 adds the cost calculator and model comparator. Phase 3 fills out the rest.