tegwick 1b01f0edf4 Honour explicit OpenRouter --model when it equals the adapter default ...

The adapter compared self._model to _DEFAULT_MODEL ("anthropic/claude-sonnet-4")
to decide whether to honour the constructor's model. When a caller passes
that exact value via --model, the comparison treats it as "not specified"
and falls through to RunConfig.model_name, which defaults to "gpt-4". So
every llm-connect call started with --provider openrouter --model
anthropic/claude-sonnet-4 actually landed on OpenAI's gpt-4 — and on
gpt-4 OpenAI's structured-output response_format requires a model with
schema support that gpt-4 lacks, returning 400. The CUST-WP-0045 canary
hit this for hours; the smoke probes that worked were the ones with no
json_schema, where gpt-4 returned fine.

Track _explicit_model separately so a constructor or LLMConfig that
matches the default is still treated as a real intent.

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

2026-06-02 14:50:37 +02:00

..

__init__.py

Implement-LLM-WP-0005-cost-model-estimators

2026-05-19 05:02:20 +02:00

_http.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00

_token_estimator.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00

adapter.py

refactor: simplify post-WP-0002 cleanup

2026-04-01 22:30:00 +00:00

claude_code.py

Prefer JSON-bearing envelope fields, skip metadata, in Claude CLI unwrap

2026-06-02 12:44:25 +02:00

cli.py

Implement-LLM-WP-0005-cost-model-estimators

2026-05-19 05:02:20 +02:00

config.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00

costs.py

Implement-LLM-WP-0005-cost-model-estimators

2026-05-19 05:02:20 +02:00

embedding_adapter.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00

embedding_cache.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00

embedding_factory.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00

embedding_openai.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00

exceptions.py

refactor: simplify post-WP-0002 cleanup

2026-04-01 22:30:00 +00:00

factory.py

feat: WP-0003 — RoutingPolicy (FR-2) and HTTP serve mode (FR-1)

2026-04-01 22:34:00 +00:00

gemini.py

refactor: simplify post-WP-0002 cleanup

2026-04-01 22:30:00 +00:00

grading.py

Add adaptive cost-quality routing primitives

2026-05-17 21:32:27 +02:00

models.py

refactor: simplify post-WP-0002 cleanup

2026-04-01 22:30:00 +00:00

openai.py

refactor: simplify post-WP-0002 cleanup

2026-04-01 22:30:00 +00:00

openrouter.py

Honour explicit OpenRouter --model when it equals the adapter default

2026-06-02 14:50:37 +02:00

problem_classes.py

Implement-LLM-WP-0005-cost-model-estimators

2026-05-19 05:02:20 +02:00

quality.py

Add adaptive cost-quality routing primitives

2026-05-17 21:32:27 +02:00

rates.py

Implement-LLM-WP-0005-cost-model-estimators

2026-05-19 05:02:20 +02:00

routing.py

Add adaptive cost-quality routing primitives

2026-05-17 21:32:27 +02:00

server.py

Preserve llm-connect run config in server mode

2026-05-19 20:55:02 +02:00

shadowing.py

Add adaptive cost-quality routing primitives

2026-05-17 21:32:27 +02:00

similarity.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00

toml_config.py

feat: initial llm-connect package scaffold

2026-02-27 07:54:42 +01:00