Claude Code stops working, user fixes it with Google's Antigravity (Gemini)
Claude Code stopped launching after a workshop installation that used a native installer, leaving the program’s directory ".local\bin" missing from the Windows PATH.
Key points
- Claude Code failed to launch after workshop install due to missing PATH entry.
- Google's Antigravity (Gemini) identified the issue and guided a reinstall via npm to restore functionality.
- Gemini also cleaned 150 legacy permission rules and helped visualize memory usage with /context.
The user turned to Google’s Antigravity IDE, powered by Gemini, which spotted the PATH trap and advised reinstalling via npm. After the reinstall, the command worked again. Gemini then helped clean up 150 legacy permission rules that had accumulated over six months, eliminating a flood of yellow warning messages.
Finally, the user used the "/context" command to visualize memory usage and experimented with folder hierarchy to optimize token consumption. The experience highlighted how an AI agent can diagnose and resolve tooling issues quickly.
The story so far
3 episodes →- Claude Code stops working, user fixes it with Google's Antigravity (Gemini)this story
[True Story] My beloved Claude Code stopped working!? How I solved the trouble and optimized the context by working hand-in-hand with Google's AI 'Antigravity (Gemini)'
note.com · 22 September 2026
Loading the full article…
This text was published by note.com and written by つくし. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Agents & Tools
All →- Meta's Muse AI agent tops US App Store, drives stock surge · 18 src
- Nvidia releases Isaac ROS 5.0 with agentic workflows and ROS Lyrical support · 2 src
- Amazon adds skill selection and instruction-following evaluators to Bedrock AgentCore · 1 src
- Meta tests Muse calls that are actually made by humans in a call center · 1 src
- Simon Willison releases llm-typesafe plugin for TypeSafe AI's Jev model · 1 src
Comments
via GitHub Discussions