[dianagrecu-uipath]

Raises the uipath floor to >=2.15.1 and bumps the version to 0.19.4.

uipath 2.15.1 contains UiPath/uipath-python#1974, which fixes the eval tool simulator for Gemini models and for reasoning models in general:

• The simulator rewrites JSON-schema type lists such as "type": ["string", "null"] into a form Gemini accepts (nullable: true, or anyOf for several types). Before this, Gemini rejected tool schemas that use type lists, for example schemas from MCP servers, with a 400 error.
• Every simulation call now sets max_tokens. With the agent's model, the value is the agent's own max tokens, capped at 16384. Otherwise it is 16384. Before this, simulation calls used the chat service default of 4096. Reasoning models count reasoning tokens against that limit, so long simulated responses could come back empty.

The only other uipath change since 2.15.0 is a documentation fix (UiPath/uipath-python#1975). uipath 2.15.1 declares the same dependencies as 2.15.0. No code in this package changes.

The upper bound stays <2.16.0. uipath 2.15.1 is on PyPI, so uv.lock is locked against it. The lock diff covers only uipath and this package's version.

Verified on the locked environment: uv lock --check passes and pytest passes (4139 passed). No source files change, so lint and type checks are unaffected.

[Copilot]

🟢 Approval recommended

No active review issues are reported for the dependency, package-version, and lockfile updates.

No open issues 🎉

What changed in this PR

Updates uipath-langchain to require the upstream eval simulator fixes for Gemini schema compatibility and reasoning-model token limits. No package source code changes.

Changes:

• Raises the uipath minimum to 2.15.1, retaining the <2.16.0 upper bound.
• Bumps uipath-langchain to 0.19.4.
• Synchronizes the lockfile with both version updates.

File	Description
uv.lock	Locks uipath to 2.15.1 and updates this package’s version and dependency metadata.
pyproject.toml	Sets version 0.19.4 and requires uipath>=2.15.1,<2.16.0.

_{Review effort: Balanced}

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud