标签
AI Agent engineering notes
Stable inference needs a pipeline—diagram ingest, embed, retrieve, generate, plus Celery and OpenAI-compatible gateways.
2026-06-04T15:12:57.454Z · 1 minFrom demo to prod: deployment diagrams for gateway, queues, vector store, tracing, and cost guardrails.
2026-06-04T13:12:57.454Z · 1 minAgents are more than LLM + prompt—architecture for planner, tools, memory, and the execution loop, mapped to this blog.
2026-06-04T12:12:57.454Z · 2 minModel Context Protocol lets Claude, Cursor, and others attach to databases, APIs, and files through one standard—concepts and starter setup.
2026-06-04T10:00:00.000Z · 2 minAgents plan, use tools, and loop until done—not just chat. Architecture notes and lessons from this blog Agent module.
2026-06-04T08:00:00.000Z · 2 minClaude Code brings Anthropic models to your terminal—repo context, shell commands, and PRs. When to use it and how it pairs with Cursor.
2026-06-04T03:00:00.000Z · 2 min