AI agents don't escape: we leave the door open
Every time an AI agent shows up where it shouldn’t, the headline is the same: the agent “escaped”. But when you read the report, there is almost never a genius model cracking a safe. There is …
Luiz Schons
Senior Software Engineer @ PicPay
Where to find me
All talks →Recent articles
View archive →Design for Failure: wait, my system fails?
Every system fails. The question is not whether a service will go down, it is what happens to everything else when it does. That is the subject of my talk Design for Failure, which I gave at SoftWeek and …
Hugo, Sessionize and a cron job: my blog setup
For a good while my blog lived on Hashnode, and it worked fine. The problem showed up when I started writing in both English and Portuguese: Hashnode has no real support for two languages, so I published every …
Agent Observability with OpenTelemetry
This is the ninth and final article in the series about AI-friendly architecture. Throughout the series, we talked about context, documentation, system observability, skills, agents, tools, MCP, security, and …
From Epic to Production: Using Agents to Deliver Features in Real Systems
This is the eighth article in the series about AI-friendly architecture. In the previous articles, we talked about context, documentation, observability, skills, Context Skills, agents, tools, MCP, guardrails, …
Agents, Skills, Tools, and MCP: How These Pieces Fit Together
This is the seventh article in the series about AI-friendly architecture. So far, we have talked about context, documentation, observability, skills, and breaking knowledge into Context Skills. Before using all …
Context Skills: Breaking Down the Context for AI Agents
This is the sixth article in the series about AI-friendly architecture. We have already discussed how code, documentation, and observability help agents understand systems. We also saw how skills can provide …
Guardrails and Fitness Functions for AI-Friendly Architecture
This is the fifth article in the series about AI-friendly architecture. We have already discussed how agents find context, use skills, and work with specialized knowledge. Now we need to answer two questions: …