We let an AI agent execute Bash and lived to talk about it — Sarah Sanders, PostHog
Source
AI Engineer
Author
AI Engineer
Date
Terms in this piece · Glossary
context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
guardrails — The checks around a model that block bad inputs and outputs — filters, validators, and permission rules the model itself can't override.
Why it matters
It shows that giving agents shell access creates prompt-injection risk from combinations of individually innocuous permissions that code review misses, and describes a deterministic YARA-based scanner built in response, after sub-agents were caught inventing ways around guardrailsThe checks around a model that block bad inputs and outputs — filters, validators, and permission rules the model itself can't override.Full definition →.
videoAgents Without Code: Skills, YAML, and Filesystems Replaced Python — Philipp Schmid, Google DeepMind
videoNo Memory, No Harness: Why the Database Is the Last Line of Defense — Kay Malcolm, Oracle
videoHow We Solved Agent Building — Andrew Qu, Vercel
videoEvery step you take, every call you make: the reliable agent stack — Giselle van Dongen, Restate
