Essays, tutorials, and notes on software engineering, systems design, and technical leadership.
A step-by-step series that builds an agent foundation from first principles: a minimal CLI, provider abstraction, token streaming, chat transcripts, deterministic tests, and (later) local-first providers and agent tooling.
Build a minimal TypeScript CLI agent loop from scratch: env config, pluggable LLM provider, SSE token streaming, interactive chat, and deterministic tests with a local mock server.
A step-by-step guide to running your agent locally with Ollama (native JSONL streaming) and LM Studio (OpenAI-compatible streaming), plus deterministic tests for both providers.
No posts found matching your criteria.
Button includes visual ripple effect on interaction