Articles for people who
ship the stack
Rewrites on React, Node.js, TypeScript and AI. A translated body is shown when it exists. If it does not, you see the English text and a note that says so.
Tagged: structured-output
Context Engineering for LLM Apps: Structured Output, Tools and Agent Limits
How to build model input deliberately: context selection, prompt templates, JSON schemas, constrained decoding, tool calling, agent loops and bounded autonomy.
13 min readBuild a Planning Agent in LangGraph: Plan, Execute, Replan, Report
A four-node LangGraph planning agent for financial research: structured plans, a namespace-aware tool registry, replanning limits and a traced example run.
16 min readMaking Two Local LLM Agents Debate: Grounding, Fake Consensus, Schema Bugs
Two local LLM agents on Ollama made to debate by evidence, not persona: NLI grounding, a fake consensus bug, attack vs assertion checks and required fields.
7 min readFrom Text Predictor to Software Component: Schemas and Function Calling
Follow how JSON Schema, structured output and constrained decoding turn an LLM's free text into typed data your code can trust, leading to function calling.
6 min readAfter the LangGraph Tutorial: Hard Edges, Schemas, and Safety Layers
Taking a LangGraph SQL agent past the tutorial: typed guards, structured output as a contract, schema refresh, cost routing and layered checks before prod.
12 min read
About these articles
Request a 24h estimate
Need the same stack in a production operator layer? Send the brief — estimate within 24 hours.