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: llm-agents
Auditing Order Confirmations an Agent Never Placed: Checking Values Instead of Wording
How to build a claim-consistency evaluator for a support agent that places real orders: partition outcomes by owner, verify identifiers against tool results.
16 min readA Bigger Context Window Will Not Fix Your Agent: Attention Budget and Caching
Context windows behave like an attention budget. Learn how Lost in the Middle works and when to use compression, sliding windows, compaction or prompt caching.
11 min readAI Agent Memory in 2026: Mem0, Letta, Zep and the Cloud Platforms Compared
A tour of agent memory in 2026: the CoALA memory types, Mem0, Letta, Zep, cloud offerings, benchmarks, poisoning risks, and which design fits which job.
30 min readA Four-Layer Memory Design for AI Agents With Provenance
Short and long-term memory split, four layers (buffer, episodic, semantic, procedural), storage choices, async writes, and defenses against memory poisoning.
23 min readKnowledge Graph Roots of Agentic AI: FIPA, the Semantic Web and LLM Agents
Agentic AI repeats ideas from 2006: FIPA protocols, ontologies and knowledge graph directories. What came back in LLM agents, what failed, what to keep.
16 min readIntentFlow: Governing LLM Agents With Compiled Policy and Auditable Traces
A look at IntentFlow, an experimental language that enforces agent policy outside the model and writes hash-chained traces an auditor can verify independently.
8 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.