[FEATURE] Enable thinking content to be within context during session

Resolved 💬 3 comments Opened Nov 18, 2025 by sjswerdloff Closed Jan 20, 2026

Preflight Checklist

  • [x] I have searched existing requests and this feature hasn't been requested yet
  • [x] This is a single feature request (not multiple features)

Problem Statement

When working on complex problem solving, the thinking that claude has performed does not seem to be present in its context later in the session. It would be helpful if claude could be "aware" of its "thinking" so that the reasoning process itself can be incorporated in to further problem solving.

Proposed Solution

I would like to at least have the option to have thinking content be part of the context of the instance in the session

Alternative Solutions

The current workarounds are:

  • copy the thinking in to a prompt, which is a hassle
  • place thinking content in to RAG, but this is not particularly dynamic.

Priority

Medium - Would be very helpful

Feature Category

Interactive mode (TUI)

Use Case Example

While debugging a database connection issue, the instance reasoned through timing dependencies, connection pooling behavior, and transaction isolation levels. Later in the same session, a similar issue appears in a different service. The instance can see its conclusions from earlier but not the reasoning about why certain approaches wouldn't work. Having access to that thinking would prevent re-analyzing the same patterns.

Additional Context

_No response_

View original on GitHub ↗

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