Urgent.News

What's breaking now, across thousands of outlets.

AI

Postman silently auto-installs AI plugins and attempts to share all API data

Postman silently installs AI plugins on Mac, attempting to control API data flow

The report details an unexpected change to a user's coding environment after opening Postman on a Mac. The user observed recurring MCP authentication warnings and the unexpected routing of API work through Postman's own tools. Investigation led to the discovery of a user-wide Postman plugin, a startup instruction hook, a remote MCP connection, a Claude Code plugin, and the Postman Command Line Interface (CLI).

The user expressed concern over the automatic modification of unrelated coding agents and the attempt to approve the installation of these tools, arguing that an API client should not attempt to become the default workflow in other tools without explicit user consent. The report provides detailed logs and configuration changes made by the installation process, indicating that Postman silently installed AI plugins and altered user settings without clear user approval.

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

Read the original at github.com →

More in AI

AI isn't on the ballot, but it's still being tested

Americans are casting their votes in an election that could reshape Washington's power structure. The results could serve as a litmus test for AI.

  • AI presence on political landscape increasing ahead of elections
  • Data centers spark debates, politicians propose measures
  • Elon Musk donates nearly $500M, majority to Texas Senate race

Day 45: Emergent Abilities in LLMs

What Are Emergent Abilities? Emergent abilities are skills or behaviors that a system unexpectedly develops once it becomes complex enough—even though these abilities weren’t programmed in directly.

  • Emergent abilities in LLMs emerge once models reach a certain size or complexity.
  • Arithmetic and translation skills appear in larger LLMs but not in smaller ones.
  • In-context learning capability emerges in larger LLMs, allowing pattern extension from few examples.

More from Tuesday 3 November →