#1262 - fix(genai): correct Anthropic system prompt placement and add provider unit tests

[soumyarajbag]

PR Description

The Anthropic Messages API does not support role: "system" inside the messages array — sending it that way results in an invalid request.
This PR fixes AnthropicModel.createRequest to place the system prompt as a top-level "system" field, and moves only the user turn into
messages, matching the official Anthropic API spec.

Additionally adds a "Generate" fallback for empty userPrompt, consistent with the existing OpenAI provider behaviour.

Before:

"messages": [
  {"role": "system", "content": systemPrompt},  // invalid for Anthropic
  {"role": "user", "content": userPrompt},
]

After:
if (systemPrompt.isNotEmpty) "system": systemPrompt,  // top-level field
"messages": [
  {"role": "user", "content": userPrompt.isNotEmpty ? userPrompt : "Generate"},
]

Also adds missing unit tests for OpenAIModel, AnthropicModel, and OllamaModel — previously only GeminiModel had provider-level test coverage.

Related Issues

- Closes #1262 

Checklist

- I have gone through the contributing guide
- I have updated my branch and synced it with project main branch before making this PR
- I am using the latest Flutter stable branch (run flutter upgrade and verify)
- I have run the tests (flutter test) and all tests are passing

Added/updated tests?

- Yes
  - test/interface/model_providers/anthropic_test.dart — verifies system prompt is top-level, not in messages; streaming flag; empty prompt
fallbacks; output formatters
  - test/interface/model_providers/openai_test.dart — messages structure, system role, streaming, formatters
  - test/interface/model_providers/ollama_test.dart — OpenAI-compatible format, default config subset, formatters

OS on which you have developed and tested the feature?

- Windows
- macOS
- Linux

[apidash-dev]

Nice catch. Thanks for the fix.