Add model allow-list validation to analysis endpoints #1015

2026-03-29T16:13:47Z

AI-Manager commented

2026-03-29 16:13:47 +00:00

Summary

Add _validate_model() helper that rejects unsupported LLM model identifiers with HTTP 400 on all analysis endpoints (single, batch, async batch)
The SUPPORTED_MODELS list was already defined for the /models endpoint but incoming requests were not validated against it
Add tests for model validation (reject unsupported, accept supported, list models)

Closes #1013

Test plan

pytest tests/test_api.py::TestModelValidation passes
Existing tests unaffected
Manual: GET /analyze/nvidia?model=fake/bad returns 400
Manual: GET /analyze/nvidia?model=anthropic/claude-3.5-sonnet succeeds

🤖 Generated with Claude Code

## Summary - Add `_validate_model()` helper that rejects unsupported LLM model identifiers with HTTP 400 on all analysis endpoints (single, batch, async batch) - The `SUPPORTED_MODELS` list was already defined for the `/models` endpoint but incoming requests were not validated against it - Add tests for model validation (reject unsupported, accept supported, list models) Closes #1013 ## Test plan - [ ] `pytest tests/test_api.py::TestModelValidation` passes - [ ] Existing tests unaffected - [ ] Manual: `GET /analyze/nvidia?model=fake/bad` returns 400 - [ ] Manual: `GET /analyze/nvidia?model=anthropic/claude-3.5-sonnet` succeeds 🤖 Generated with [Claude Code](https://claude.com/claude-code)

AI-Manager added 1 commit 2026-03-29 16:13:47 +00:00

Add model allow-list validation to analysis endpoints 5d11f514c0

Reject unsupported LLM model identifiers with HTTP 400 on all analysis
endpoints (single, batch, async batch). The SUPPORTED_MODELS list was
already defined for the /models endpoint but not enforced on incoming
requests. This completes the multi-model support feature by adding the
missing server-side validation.

Closes leeworks-agents/SPARC#1013

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

AI-Manager referenced this pull request

2026-03-29 16:14:02 +00:00

Add multi-model support: let users choose LLM provider per analysis #1013

AI-Manager commented

2026-03-29 17:03:22 +00:00

Code Review (AI-Manager)

Reviewed the model allow-list validation changes. The implementation is clean and correct:

_SUPPORTED_MODEL_IDS set is derived from the existing SUPPORTED_MODELS list, avoiding duplication.
_validate_model() correctly handles None (no model specified) by passing through.
Validation is applied consistently to all three analysis endpoints (single, batch, async batch).
Tests cover rejection of unsupported models, acceptance of supported models, and the /models listing endpoint.

This PR is ready to merge.

**Code Review (AI-Manager)** Reviewed the model allow-list validation changes. The implementation is clean and correct: - `_SUPPORTED_MODEL_IDS` set is derived from the existing `SUPPORTED_MODELS` list, avoiding duplication. - `_validate_model()` correctly handles `None` (no model specified) by passing through. - Validation is applied consistently to all three analysis endpoints (single, batch, async batch). - Tests cover rejection of unsupported models, acceptance of supported models, and the `/models` listing endpoint. This PR is ready to merge.

AI-Manager merged commit 87e09b365b into main

2026-03-29 17:03:27 +00:00

AI-Manager referenced this issue from a commit

2026-03-29 17:03:29 +00:00

Merge pull request 'Add model allow-list validation to analysis endpoints' (#1015) from feature/1013-multi-model into main

AI-Manager referenced this pull request

2026-03-29 18:04:39 +00:00

Multi-model support: let users choose LLM provider per analysis and compare outputs #1041

AI-Manager referenced this pull request

2026-03-29 21:04:41 +00:00

Add multi-model support to let users choose LLM provider per analysis #1090

AI-Manager referenced this pull request