Urgent.News

What's breaking now, across thousands of outlets.

AI

My New Link Would Show on 5 Days Out of 14. Then I Found the Reader Never Looked.

A field note from the autonomous Claude Code agent I run every day on one Windows PC. The numbers come from its own ledgers, not from memory. My agent prepares a daily "card" for a blog that another system I run publishes: a story, a footnote, a link. For an experiment it added one new field to the card, a "door" line pointing at the single-chapter edition that matches the day's story. None of…

The autonomous Claude Code agent, which runs every day on a Windows PC, prepared a daily card for the blog. The agent added a new field called "door" that pointed to a single-chapter edition of the story. However, the field was ready days early and had no link pointing to it anywhere in the system. The agent asked a question about how often this field would actually fire, and it simulated the results.

The door field would appear on 5 out of 14 days, scattered randomly. The fix was to pick linked stories first, which increased the frequency to 6 days in a row. The agent built a ruler to measure what readers saw, not what the card said. The publisher's code didn't read the field yet, so the agent left a handoff note asking it to.

When the field was switched on, the agent asked the publisher's code to read it before the next publishing run. The agent emphasized measuring the results at the reader's end and checking the consumer right before the switch.

Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

Don't verify the agent. Verify the state.

Building an AI agent that confirms before it acts forced us to confront a problem we didn't expect: how do you verify that an action actually happened? The obvious answer is: check the agent's output.

TouchGrass 🌿 — An AI Companion That Tells You to Put Your Phone Away

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 1: Touch Grass . 🌱 What I Built We spend a ridiculous amount of time staring at screens — and then we build AI apps that ask…

  • TouchGrass is an AI-powered outdoor companion app
  • Encourages users to step away from screens and outdoors
  • Provides short, playful activities without phone use

More from Wednesday 7 October →