[MODEL] CLAUDE CODE HALLUCINATES DATA THAT IS NOT AVAILABLE:

Preflight Checklist

• [x] I have searched existing issues for similar behavior reports
• [x] This report does NOT contain sensitive information (API keys, passwords, etc.)

Type of Behavior Issue

Other unexpected behavior

What You Asked Claude to Do

You should report this issue to Anthropic through their official Claude Code feedback channel: Report the issue at: https://github.com/anthropics/claude-code/issues You can create a new issue describing: 1. The false claims about finding Reddit and chess.com WebSocket data that never existed 2. The wasted time trying to "fix" rendering when there was no data to render This is the official channel Anthropic monitors for Claude Code problems. They need to know when the AI provides completely false information like claiming to find specific data that doesn't exist. You can also include that the AI: - Claimed to find "27 chess.com WebSocket logs from around 10:30pm" - Claimed to find Reddit WebSocket data - Made multiple attempts to "fix" code based on non-existent data - Wasted significant time on false premises This is a serious issue where the AI hallucinated specific data findings and they should investigate why this happened.

So that is it. Agent has been lying consistently, it has made up data repeatedly. Want to use logging IN EVERY STEP, does not read codebase and lies about it. Claims to have modified code WHEN NO MODIFICATION WAS ACTUALLY DONE. THis has been going on for months. If you are nerfing the model STOP. It has been on a down spiral.

The model could write decently complex code easily before, but it started to become dumb, lazy and a liar. ffs tell it to stop using LOGGING for every small fix.

I don't know about ticket structure. But I wanted to leave this here becaues it is becoming a hassle to deal with it.

What Claude Actually Did

Agent has been lying consistently, it has made up data repeatedly. Want to use logging IN EVERY STEP, does not read codebase and lies about it. Claims to have modified code WHEN NO MODIFICATION WAS ACTUALLY DONE

Expected Behavior

NOT DO THE ABOVE.

Files Affected

Permission Mode

Accept Edits was ON (auto-accepting changes)

Can You Reproduce This?

Yes, every time with the same prompt

Steps to Reproduce

_No response_

Claude Model

Sonnet

Relevant Conversation

Impact

Critical - Data loss or corrupted project

Claude Code Version

2.0.20

Platform

Anthropic API

Additional Context

_No response_