agent: Fix issue with Anthropic thinking models (#33317)

cc @osyvokon 

We were seeing a bunch of errors in our backend when people were using
Claude models with thinking enabled.

In the logs we would see
> an error occurred while interacting with the Anthropic API:
invalid_request_error: messages.x.content.0.type: Expected `thinking` or
`redacted_thinking`, but found `text`. When `thinking` is enabled, a
final `assistant` message must start with a thinking block (preceeding
the lastmost set of `tool_use` and `tool_result` blocks). We recommend
you include thinking blocks from previous turns. To avoid this
requirement, disable `thinking`. Please consult our documentation at
https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking

However, this issue did not occur frequently and was not easily
reproducible. Turns out it was triggered by us not correctly handling
[Redacted Thinking
Blocks](https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking#thinking-redaction).

I could constantly reproduce this issue by including this magic string:
`ANTHROPIC_MAGIC_STRING_TRIGGER_REDACTED_THINKING_46C9A13E193C177646C7398A98432ECCCE4C1253D5E2D82641AC0E52CC2876CB
` in the request, which forces `claude-3-7-sonnet` to emit redacted
thinking blocks (confusingly the magic string does not seem to be
working for `claude-sonnet-4`). As soon as we hit a tool call Anthropic
would return an error.

Thanks to @osyvokon for pointing me in the right direction 😄!


Release Notes:

- agent: Fixed an issue where Anthropic models would sometimes return an
error when thinking was enabled
This commit is contained in:
Bennet Bo Fenner 2025-06-24 18:23:59 +02:00 committed by GitHub
parent 39dc4b9040
commit 7be57baef0
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
7 changed files with 36 additions and 10 deletions

View file

@ -1030,6 +1030,7 @@ pub fn response_events_to_markdown(
Ok(LanguageModelCompletionEvent::Thinking { text, .. }) => {
thinking_buffer.push_str(text);
}
Ok(LanguageModelCompletionEvent::RedactedThinking { .. }) => {}
Ok(LanguageModelCompletionEvent::Stop(reason)) => {
flush_buffers(&mut response, &mut text_buffer, &mut thinking_buffer);
response.push_str(&format!("**Stop**: {:?}\n\n", reason));
@ -1126,6 +1127,7 @@ impl ThreadDialog {
// Skip these
Ok(LanguageModelCompletionEvent::UsageUpdate(_))
| Ok(LanguageModelCompletionEvent::RedactedThinking { .. })
| Ok(LanguageModelCompletionEvent::StatusUpdate { .. })
| Ok(LanguageModelCompletionEvent::StartMessage { .. })
| Ok(LanguageModelCompletionEvent::Stop(_)) => {}