Skip to content

Add --output-mode flag to filter output#63

Open
czue wants to merge 2 commits intosimonw:mainfrom
czue:cz/filter-output
Open

Add --output-mode flag to filter output#63
czue wants to merge 2 commits intosimonw:mainfrom
czue:cz/filter-output

Conversation

@czue
Copy link

@czue czue commented Feb 5, 2026

Really like this tool!

I wanted to be able to share transcripts without leaking the tool call results (which contained proprietary data/code). This adds an --output-mode flag (documented in the README) to allow different modalities.

You can see how it works by viewing these three logs of the conversation I used to have Claude write the code:

  • Full output - This is the same as before and the default.
  • Compact mode - Includes tool calls but not tool results (my use case)
  • Conversation mode - Just includes the transcript of user and Claude's messages

@simonw would love your input on this. This also addresses (at least partially) #61

czue and others added 2 commits February 5, 2026 13:36
Adds OutputFilter dataclass with three presets:
- full (default): show everything
- compact: hide tool results (useful for hiding sensitive output)
- conversation: show only user/assistant text (no tools, thinking)

The backend uses individual boolean flags (hide_tool_calls,
hide_tool_results, hide_thinking) with a should_hide() method,
making it easy to add granular CLI flags later.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant