fix: preserve thought_signature across tool-call turns (fixes Gemini 3.x tool use)

[FritzHeider]

Gemini 3.x rejects a replayed function call whose thought_signature is missing, so the agent dies on its second step:

400 INVALID_ARGUMENT: Function call is missing a thought_signature in
functionCall parts. This is required for tools to work correctly.
Additional data, function call `default_api:goto_tool`, position 2.

Step 1 succeeds (the browser navigates), then every retry fails and the run aborts. This makes the agent unusable on current Gemini models — gemini-2.5-flash and gemini-2.0-flash now return 404 ... no longer available to new users for new API keys, so new users land on a 3.x model by default.

Root cause

The Google provider already handles both ends of this:

• extracts the signature from the response — src/providers/google/llm.py:321 (_extract_thought_signature, into Thinking(signature=...))
• replays it onto the function-call Part when rebuilding history — src/providers/google/llm.py:121-126

The gap is between them. src/agent/service.py built its ToolMessage from the LLMEvent without thinking or thinking_signature, so the extracted signature was discarded before it could ever be replayed. ToolMessage.thinking_signature is declared in src/messages/service.py:150 but nothing assigned it anywhere in the codebase.

Fix

Thread both fields from the LLMEvent into the ToolMessage (2 lines).

Also affects Anthropic

src/providers/anthropic/llm.py:148,160 reads the same msg.thinking_signature that nothing previously set, so extended thinking with tools was likely broken in the same way. This fix is provider-agnostic and should repair that path too — though I verified on Gemini only.

Verification

With gemini-flash-latest:

• before: fails at step 2 with the error above, all 3 retries exhausted
• after: a multi-step navigate-and-extract task completes and returns its final answer

Happy to add a regression test if you would like one — a fake provider returning a signature and asserting it survives the history round-trip would cover it.

[qodo-code-review]

PR Summary by Qodo

Fix: preserve thought_signature when recording tool-call turns

🐞 Bug fix 🕐 10-20 Minutes

AI Description

• Propagate LLM thinking and thought_signature into ToolMessage tool-call turns.
• Prevent Gemini 3.x tool-call replay failures caused by missing thought_signature.
• Unblocks Anthropic extended-thinking tool use by preserving the same signature field.

Diagram

graph TD
A["Agent loop (service.py)"] --> B["LLM provider ainvoke()"] --> C["LLMEvent (thinking+signature)"] --> D["ToolMessage history"] --> E["Provider _convert_messages()"] --> F{{"Gemini/Anthropic API"}}

Loading

High-Level Assessment

This is the right minimal fix: the providers already extract and replay thought_signature, so the correct place to preserve it is when the agent records the tool-call turn (ToolMessage). The only meaningful follow-up to consider is adding a regression test that asserts thinking_signature survives an LLMEvent → ToolMessage → provider history rebuild round-trip.

Files changed (1) +2 / -0

Bug fix (1) +2 / -0

service.pyPreserve thinking and thought_signature in ToolMessage tool-call turns +2/-0

Preserve thinking and thought_signature in ToolMessage tool-call turns

• When the LLM returns a TOOL_CALL event with Thinking metadata, the agent now copies both thinking content and thinking_signature into the ToolMessage it records. This prevents signature loss across turns so providers can correctly replay signed tool calls (required by Gemini 3.x; also used by Anthropic extended thinking).

src/agent/service.py

[qodo-code-review]

Code Review by Qodo

🐞 Bugs (0) 📘 Rule violations (0) 📎 Requirement gaps (0)

Great, no issues found!

Qodo reviewed your code and found no material issues that require review

_{To customize comments, go to the Qodo configuration screen, or learn more in the docs.}