Cómo probar agentes de IA contra fallos de producción antes que tus usuarios
💻 Este es el inicio de una serie. Todo el código está en un solo repo:...
All Posts
-
-
Detener alucinaciones de agentes de IA: Validar antes de que el agente escriba en memoria
💻 Todo el código de esta serie está en un solo repo: resilient-agent-harness-sample-for-aws. Este...
-
Fable 5 returns, agent memory levels up. Tech & AI Weekly by Eli, July 3 2026
Claude Fable 5 returns to Bedrock, Sonnet 5 lands on AWS, and AgentCore memory gets metadata filtering. This week in tech and AI.
-
A fake bug report can hijack your coding agent. Tech & AI Weekly by Eli, June 26 2026
This week: AgentJacking hijacks Claude Code, Cursor, and Codex through a fake Sentry bug report, new research on why models confuse trusted and untrusted text, Strands chaos testing, AWS Blocks, and Lambda MicroVMs.
-
How to Stop Prompt Injection in AI Agents That Read Untrusted Content
When an AI agent reads untrusted content (a web page, a document, an email), a hidden instruction can ride in, get stored in the agent's own memory, and fire in
-
How to Test AI Agents for Production Failures Before Your Users Do
An AI agent that's flawless in the demo can still fall apart the first time a tool fails in production: a timeout, a network error, a response that comes back c
-
Self-Improving AI Agents: Turn Repeated Reasoning Into Tools the Agent Writes Itself
A static AI agent re-reasons the same kind of task from scratch every time, burning tokens and sometimes getting it wrong differently on each run. A self-improv
-
Stop AI Agent Hallucinations: Validate Before the Agent Writes to Memory
When an AI agent hallucinates a fact, the real damage starts when it writes that fact to memory and re-reads it as trusted context every session after, compound
-
Why AI Agents Fail at Multi-Step Tasks, and How to Catch the Silent Failure
On a multi-step task, an AI agent will trust a tool that reports success even when the work silently never saved, and then confidently report the whole task don
-
Strands cut agent costs in half, and I published a book — Tech & AI Weekly, June 19 2026
This week: I published my book on AI agents that don't hallucinate or waste tokens, Strands cut agent costs in half, GLM-5.2 open weights, and the agent launches from AWS Summit New York.
