fix(ollama): add Gemma to tool support heuristic for Ollama and LM Studio

[octo-patch]

Fixes #12131

Problem

Gemma 3/4 models running via Ollama or LM Studio are not recognized as supporting tool calling, so Continue never offers tools to them. This affects users who run Gemma models locally through either provider.

The root cause: the Ollama provider's static model heuristic in toolSupport.ts is missing "gemma". The LM Studio provider delegates to Ollama's heuristic, so both are affected.

Without this entry, modelSupportsNativeTools() returns false for any Gemma model, and tools are never offered or sent — even if the model's Ollama template includes .Tools support (the secondary check added in #11670).

Note: the openai provider already lists Gemma as supported, and the existing test suite confirms this (lines 109-113 in toolSupport.test.ts).

Solution

Add "gemma" to the Ollama supported-models list with a reference to Google's function calling documentation:
https://ai.google.dev/gemma/docs/capabilities/function-calling

For Ollama, the /api/show template check (from #11670) will still act as a secondary gate — if the specific Gemma model's template doesn't include .Tools, tools will be skipped. This PR only removes the pre-filter that was blocking Gemma entirely.

Testing

Added test cases for Gemma models in both ollama and lmstudio provider sections. All 83 tests pass.

Tests: 83 passed, 83 total

---

Summary by cubic

Enable native tool calling for Gemma 3/4 models when run via ollama or lmstudio. Adds "gemma" to the Ollama tool-support heuristic so Continue offers tools; LM Studio inherits this, and the existing /api/show template check still gates per-model support.

^{Written for commit 990aa9e. Summary will update on new commits.}

[github-actions]

Thank you for your submission, we really appreciate it. Like many open-source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution. You can sign the CLA by just posting a Pull Request Comment same as the below format.

---

I have read the CLA Document and I hereby sign the CLA

---

octo-patch seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You can retrigger this bot by commenting recheck in this Pull Request. Posted by the CLA Assistant Lite bot.}

[cubic-dev-ai]

No issues found across 2 files