agentjido / req_llm / c7e7f574e42f0e1c4d01111a8e3c956f2c8f5f7f-PR-151

53%

main: 49%

Build Type

Pull #151

github

Committed by neilberkman

Commit Message

Add Mistral AI model support and fix model sync patch merging

Add comprehensive Mistral AI support to AWS Bedrock provider with 5 models:
- mistral.mistral-7b-instruct-v0:2
- mistral.mixtral-8x7b-instruct-v0:1
- mistral.mistral-small-2402-v1:0
- mistral.mistral-large-2402-v1:0
- us.mistral.pixtral-large-2502-v1:0

Architecture improvements:
- New Mistral formatter wraps Converse API with system message workaround
- Callback system for model-specific capabilities (requires_converse_api?,
  supports_converse_tool_choice?, preserve_inference_profile?)
- Inference profile preservation for region-specific model IDs

Fix model sync task to properly merge patches from priv/models_local/ for
providers with hyphenated IDs (e.g., amazon-bedrock). The task now correctly
handles provider ID normalization when looking up and merging with upstream
models.dev data.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

Pull Request Pull Request #151: Add Mistral AI model support and fix model sync patch merging

Run Details

40 of 96 new or added lines in 5 files covered. (41.67%)

11 existing lines in 3 files now uncovered.

3762 of 7143 relevant lines covered (52.67%)

224.81 hits per line