Skip to content

fix(ai-client): prevent infinite tool call loop when finishReason is stop #412

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
AlemTuzlak merged 2 commits into from
Apr 1, 2026
+15 −1
Merged

fix(ai-client): prevent infinite tool call loop when finishReason is stop #412

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
ccf67f8
fix(ai-client): prevent infinite tool call loop when finishReason is …
AlemTuzlak Apr 1, 2026
1b5c400
chore: add changeset for infinite tool loop fix
AlemTuzlak Apr 1, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 7 additions & 0 deletions .changeset/fix-infinite-tool-loop.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,7 @@
---
'@tanstack/ai-client': patch
---

fix: prevent infinite tool call loop when server tool finishes with stop

When the server-side agent loop executes a tool and the model finishes with `finishReason: 'stop'`, the client no longer auto-sends another request. Previously this caused infinite loops with non-OpenAI providers that respond minimally after tool execution.
9 changes: 8 additions & 1 deletion packages/typescript/ai-client/src/chat-client.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -657,11 +657,18 @@ export class ChatClient {
await this.drainPostStreamActions()

// Continue conversation if the stream ended with a tool result (server tool completed)
// but ONLY if the model indicated it wants to continue (finishReason !== 'stop').
// When finishReason is 'stop', the model is done — don't re-send.
if (streamCompletedSuccessfully) {
const messages = this.processor.getMessages()
const lastPart = messages.at(-1)?.parts.at(-1)
const { finishReason } = this.processor.getState()

if (lastPart?.type === 'tool-result' && this.shouldAutoSend()) {
if (
lastPart?.type === 'tool-result' &&
finishReason !== 'stop' &&
this.shouldAutoSend()
) {
try {
await this.checkForContinuation()
} catch (error) {
Expand Down
Loading