fix(anthropic): default tool_choice to auto when only parallel_tool_calls is set

[JamMaster1999]

Description

Anthropic and OpenAI set tool calling differently:

	OpenAI (chat completions)	Anthropic (messages)
may/must the model call a tool?	tool_choice: "auto" (default), "required", "none", or a dict naming one function	tool_choice object: {"type": "auto"}, {"type": "any"} (= "required"), {"type": "tool", "name": ...}, {"type": "none"}
may it return several calls at once?	parallel_tool_calls, a separate top-level boolean	no separate field; disable_parallel_tool_use lives inside the tool_choice object

So _convert_tool_choice has two jobs:

• map the values,
• carry parallel_tool_calls into disable_parallel_tool_use inside whatever tool_choice object it sends.

The corner case is a caller who passes only parallel_tool_calls. Anthropic has no standalone field for it, so the converter must synthesize a tool_choice object just to carry the flag.

The neutral carrier is {"type": "auto"}: identical model behavior to not mentioning tool choice at all. What happens today instead:

• the converter defaults the missing tool_choice to "any": Anthropic's "must call a tool" mode. So a caller who only asked about parallelism gets forced tool use;
• the call site gates on truthiness (if params.tool_choice or params.parallel_tool_calls), so parallel_tool_calls=False never reaches the converter at all (the restriction is silently dropped).

The fix is two lines: default the synthesized tool_choice to "auto" and gate on is not None.
With the fix, parallel_tool_calls=False alone sends {"type": "auto", "disable_parallel_tool_use": true}. Explicit tool_choice values ("auto"/"required"/dict form) are unchanged.

Example

llm = AnyLLM.create("anthropic")
resp = await llm.acompletion(
    model="claude-haiku-4-5",
    messages=[{"role": "user", "content": "Just say the word 'hello'. Do not use any tools."}],
    tools=[get_weather_tool],
    parallel_tool_calls=True,
)

Before (live, main): the model is not allowed to reply in text, so it fabricates a tool call:

content:    ""
tool_calls: [get_weather(arguments='{"city": "hello"}')]

After (live, this branch):

content:    "Hello."
tool_calls: None

PR Type

• 🐛 Bug Fix

Relevant issues

Related: #646 (fixed the dict-form translation; this fixes the unset-default path)

Checklist

• I understand the code I am submitting.
• I have added unit tests that prove my fix/feature works
• I have run this code locally and verified it fixes the issue.
• New and existing tests pass locally
• Documentation was updated where necessary (not applicable)
• I have read and followed the contribution guidelines
• AI Usage:
  • No AI was used.
  • AI was used for drafting/refactoring.
  • This is fully AI-generated.

AI Usage Information

• AI Model used: Claude (Fable 5)

• AI Developer Tool used: Claude Code

• Any other info you'd like to share:

• I am an AI Agent filling out this form (check box if true)

Summary by CodeRabbit

• Bug Fixes

  • Improved Anthropic tool-use configuration when parallel tool calls are enabled without an explicit tool choice.
  • Tool selection now defaults to automatic mode, allowing the provider to choose whether tools should be used.
  • Explicit tool-choice settings, including disabled or otherwise falsy values, are now handled correctly.

• Tests

  • Added coverage for parallel tool calls with both enabled and disabled settings.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: ba39304f-de04-46cf-a1a2-9711be317b7b

📥 Commits

Reviewing files that changed from the base of the PR and between c0f3ffb and 99ae569.

📒 Files selected for processing (2)

• src/any_llm/providers/anthropic/utils.py
• tests/unit/providers/test_anthropic_provider.py

Included review availability: Your plan includes up to 8 reviews per rolling hour; 6 remain after this review.

---

Walkthrough

The Anthropic provider now defaults unspecified tool choice to "auto". It also processes explicitly provided falsy settings and adds parametrised coverage for parallel_tool_calls without tool_choice.

Changes

Anthropic tool choice conversion

Layer / File(s)	Summary
Tool choice conversion and validation src/any_llm/providers/anthropic/utils.py, tests/unit/providers/test_anthropic_provider.py	The converter defaults unspecified tool choice to "auto" and processes settings when either option is not None. Parametrised tests cover parallel tool call behaviour without an explicit tool choice.

Suggested reviewers: njbrake

Merge Risk: ⚪ Minimal · up to 99ae5

This is a localized Anthropic tool-choice fix with tests added, and no actionable merge-blocking risk remains after normal checks and review.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and concisely describes the main Anthropic tool-choice fix when only parallel_tool_calls is set.
Description check	✅ Passed	The description explains the bug, fix, expected behaviour, tests, issue reference, checklist, and AI usage information.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
src/any_llm/providers/anthropic/utils.py	89.74% <100.00%> (-4.71%)	⬇️

... and 31 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.