MCP is useful, but most of the time you do not actually need it. It gives an agent a clean way to...
MCP vs Skills: Why Skills Save Context Tokens
Related
Building Agentic Workflows in Python
When a tool-calling loop earns its complexity over a single call or a scripted pipeline, the manual and SDK-provided agentic loops in Python, and the guardrails — validation, appro...
Building Reliable LLM Applications in Java
Typed structured outputs, retries and idempotency, grounding over hallucination, evaluation, prompt caching, and agentic control flow — best practices for LLM apps on the JVM with ...
My credential rule reported 842 secrets in vercel/ai. The real count was 0.
Our no-hardcoded-credentials rule fired 842 times on vercel/ai. The peer plugin fired 380. I assumed we had better recall — until I sampled. 807 of the 'extra' findings were TypeSc...