[Question] Is automatic tool calling output truncation supported? #4826

Activity

1. github-actions commented on Nov 27, 2025

   

   github-actionsbot

   on Nov 27, 2025 – with GitHub Actions

   Contributor

   More actions

   This issue might be a duplicate of existing issues. Please check:

   • [FEATURE] Sliding window context management for long-running sessions #4659: [FEATURE] Sliding window context management for long-running sessions - Addresses token consumption and performance in long sessions
   • Model in BUILD mode does not have enough context to continue after compaction #3031: Model in BUILD mode does not have enough context to continue after compaction - Related context and token management issue
   • exclude node_modules #4733: exclude node_modules - Related concern about large outputs consuming tokens unnecessarily

   Feel free to ignore if your specific case about automatic tool output truncation differs from these context management discussions.
2. Mushoz commented on Nov 27, 2025

   

   Mushoz

   on Nov 27, 2025

   Author

   More actions

   This one is a little bit extreme, but maybe it explains the issue I am having:

   {"error":{"code":400,"message":"the request exceeds the available context size, try increasing it","type":"exceed_context_size_error","n_prompt_tokens":1463244,"n_ctx":262144}}
   

   It was caused by the model trying to grep something with way too many results
3. nikhilmaddirala commented on Nov 27, 2025

   

   nikhilmaddirala

   on Nov 27, 2025

   More actions

   Claude Code handles this quite well by delegating such tasks to the "explore" subagent so the tool calls don't pollute the main agent's context window. I wonder if you try defining such an agent for yourself. Or even just instructing the model to use Opencode's "general" subagent to delegate such tasks.
4. rekram1-node commented on Nov 27, 2025

   

   rekram1-node

   on Nov 27, 2025

   Collaborator

   More actions

   we can be fancier here but rn we do automatically prune sufficiently old tool calls from context to free up space
5. Mushoz commented on Dec 9, 2025

   

   Mushoz

   on Dec 9, 2025

   Author

   More actions

   Claude Code handles this quite well by delegating such tasks to the "explore" subagent so the tool calls don't pollute the main agent's context window.

   That wouldn't fix this particular case though. This single tool call returned over a million token. A subagent would fail to run with such a big tool call result as well. Sure, it won't pollute the main agent's context window, but it's not retrieving the right answer either.

   we can be fancier here but rn we do automatically prune sufficiently old tool calls from context to free up space

   This is only helpful if the tool call itself did not result in the context window overflowing. And even if it didn't, it's still a lot of trash in the context. Would it be a bit think to just truncate the output after 500 lines or so (maybe make it configurable). And if it is truncated, let the model know with some text like "Truncated, only showing first 500 lines out of 21416".
6. thdxr

   closed this as completedon Mar 25, 2026