chore: bump camel-ai to 0.2.91a5

[fengju0213]

Related Issue

Closes #1711

Description

Bump camel-ai[eigent] from 0.2.91a3 to 0.2.91a5 and refresh the backend lockfile.

CAMEL 0.2.91a5 includes camel-ai/camel#4171, which keeps all tools available but disables strict input grammar constraints when an Anthropic request exceeds the provider's request-wide structured-output limits. This prevents Sonnet and Fable requests routed through LiteLLM from failing with Too many strict tools before inference starts.

Testing Evidence (REQUIRED)

• uvx --from uv==0.11.28 uv lock --check

  • Passed; 208 packages resolved.

• .venv/bin/pytest -q tests/app/model/test_chat.py tests/app/controller/test_model_controller.py

  • 38 passed in 0.99s

• Installed the frozen backend environment and confirmed camel-ai 0.2.91a5 loads from the PyPI wheel.

• Replayed the two real 49-tool payloads that previously failed:

  • claude-sonnet-5: 49 tools retained, 0 tools left strict, original input unchanged.
  • claude-fable-5: 49 tools retained, 0 tools left strict, original input unchanged.

• I have included human-verified testing evidence in this PR.

• This PR includes frontend/UI changes, and I attached screenshot(s) or screen recording(s).

• No frontend/UI changes in this PR.

What is the purpose of this pull request?

• Bug fix
• New Feature
• Documentation update
• Other

Contribution Guidelines Acknowledgement

• I have read and agree to the Eigent Contribution Guideline

[YunfeiZHAO]

LGTM