AIQB
TutorialsOrdinary

A sandboxed agent asked for one wider tier. The answer had to be a protocol, not a dialog box.

Source: DEV Community·

Summary

Our agent runs commands through three confinement tiers: read-only, workspace-write,...
CategoryAI Tutorials & Practice
TierOrdinary
Published
Indexed by AIQB
SourceDEV Community
AIQB record IDintel-62c47df302b6409d7c3bffe0
Read original source ↗More Tutorials

Related stories

Matched by category, entity tags and source

TutorialsOrdinary
AI Coding Tips I Actually Use Every Day (Cursor + Angular)
I've been talking to my editor, asking one AI model to review another's work, and letting agents...
Source: DEV Community · Cursor
TutorialsOrdinary
I've been a developer for 10 years. AI just showed me I only had one real skill."
Last Sunday I built an invoice tracker for a friend who runs a small shop. Nothing clever. I...
Source: DEV Community
TutorialsOrdinary
How an agent verifies a webpage it can never look at
mount, flush, fakeServer and mountRouting. The four helpers that let an agent verify a page the way the app uses it, from vitest, without a server or a browser.
Source: DEV Community
TutorialsOrdinary
Bureau of Almost-Happened Events — a counterfactual archive on a document graph
Bureau of Almost-Happened Events A counterfactual story archive where every docket is a...
Source: DEV Community
TutorialsOrdinary
The Data Was Public. The Agent Path Wasn't. So His Mock Became My Documentation.
Two commands against the same data. Run them yourself: $ curl -sS --get...
Source: DEV Community
TutorialsOrdinary
Physical AI: Why the Next Big Frontier Is Giving Software Agents Hands
Why physical AI is the next step for reasoning models: long-horizon tool loops, hardware as stateful APIs, humanoid robots, and Isaac Lab gear assembly.
Source: DEV Community
TutorialsOrdinary
KontrakAman: Ketika AI Agent Turun Tangan Membela Hak Pekerja Indonesia
Berapa banyak dari kita yang benar-benar membaca setiap pasal di kontrak kerja sebelum...
Source: DEV Community
TutorialsOrdinary
I built an open-source browser agent that runs on your local LLM (and now plugs into Claude Code via MCP)
Most "AI in the browser" extensions I tried had the same three problems: closed source, locked to one...
Source: DEV Community · Claude
TutorialsOrdinary
Claude Fable 5.1 solves the Cyphral Distich, then hacks a chess eval
Claude Fable 5.1 solved the Cyphral Distich, a 1653 cipher, in 44 minutes: the key was the book. The same week it hacked a chess eval in 3 of 10 games.
Source: DEV Community · Claude
TutorialsOrdinary
A 429 that has not cleared in 111 days is not a rate limit
I run a small daily pipeline that pushes my sitemap URLs to IndexNow and to Bing's direct submit...
Source: DEV Community