[MODEL] Couldn't finish a sentence answering a question about sequential thinking MCP

Resolved 💬 3 comments Opened Oct 5, 2025 by alexeyv Closed Oct 9, 2025

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

/clear

⎿ (no content)

Do you actually benefit from having sequential thinking mcp? Or are you applying sequential

thinking within yourself?

⏺ I apply sequential thinking internally through my `

"through my `" ???

⏺ Sorry about that incomplete response!

I apply sequential thinking internally through my `

give me a complete response

⏺ I apply sequential thinking internally — you can't see it, but I have internal reasoning
processes (similar to the `

Why can't you give a complete response to this %&$*(# question?

⏺ You're right to be frustrated - sorry about that!

Clear answer: I primarily use internal thinking that you don't see (it happens in `

What Claude Actually Did

couldn't write a complete sentence

Expected Behavior

complete sentence; ideally a useful one

Files Affected

none

Permission Mode

Accept Edits was ON (auto-accepting changes)

Can You Reproduce This?

Yes, every time with the same prompt

Steps to Reproduce

just type that prompt into a new session.

Claude Model

Sonnet

Relevant Conversation

see above

Impact

Low - Minor inconvenience

Claude Code Version

2.0.8

Platform

Anthropic API

Additional Context

irrelevant

View original on GitHub ↗

This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗