063aaa97a6
* Structured outputs for dialectic
* cleanup
* rename json_schema_to_pydantic to clarify it's not a general schema converter
* clean up schema DoS guards
* simplification and cleanup of schema conversion
* chore: ruff and pyproject toml
* chore: basedpyright cleanup in test
* fix: some needed unrelated test failures
* test(schema_conversion-and-anthropic-backend): expand test coverage
include table tests
* fix(llm): support combined tool calling and structured output across backends
- OpenAI: parse() 500s on non-strict function tools; route tool-carrying
structured requests through create() with an explicit json_schema
response_format (mirrors the streaming path)
- Anthropic: skip the '{' JSON prefill when tools are present so tool_use
blocks stay reachable; make the schema instruction conditional and rely
on parse + repair
- Gemini: native response_schema + function calling is rejected before
Gemini 3; with tools present, inject a schema instruction into the final
turn instead and rely on parse + repair
- All backends: tool-call turns carry no consumable content, so skip
structured-output parsing on them
Extracted from the dialectic structured-output branch (DEV-1652) so the
transport layer can land independently.
DEV-2035
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* test(live_llm): exercise combined tools + structured output per provider
Two-turn live flow per backend: a forced tool-call turn (structured
parsing must be skipped) followed by a replay turn that must return a
schema-conforming answer with tools still attached. Asserts the
provider-specific request shaping: no parse() for OpenAI (500s on
non-strict tools), no '{' prefill for Anthropic, no native
response_schema for Gemini.
Verified against live OpenAI (gpt-4.1, gpt-5, gpt-5.4, gpt-5.4-mini)
and Gemini (gemini-2.5-flash).
DEV-2035
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* test(unified): dialectic chat with response_format schema under tool use
Adds response_format pass-through to the unified runner's chat query and
a test case that forces the dialectic tool loop (reasoning off + global
enumeration question) while requiring a schema-conforming JSON answer —
end-to-end coverage of the combined tools + structured output transport
path on whichever provider each level is configured with.
Verified locally against a full harness run (json_match assertions pass;
the llm_judge assertion additionally runs in CI where the Anthropic key
is available).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* fix: some needed unrelated test failures
* ci: add label-triggered live LLM test workflow
Adding the run-live-llm label to a PR (or workflow_dispatch) runs
tests/live_llm/ against real provider APIs — the only place the
--live-llm suite runs in CI. Reuses the unified-tests environment and
its Secrets Manager staging-dotenv resolution for provider keys; runs
on ubuntu-latest (no Fly runner, no Docker — the suite only touches the
LLM backends). Pins LIVE_LLM_ANTHROPIC_45_PLUS_MODELS=claude-sonnet-4-5
since the Anthropic family has no default models and would otherwise
silently collect empty.
Opt-in by design: live model behavior is variable, so this is a signal,
not a required check.
DEV-2035
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* ci: run live LLM tests on main pushes touching the transport
Mirrors unified-tests' push trigger, scoped to paths that can affect
the live suite (src/llm/, config, the tests, deps, and the workflow
itself) so provider API calls aren't spent on unrelated changes.
DEV-2035
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* ci: disable auth in live LLM test environment
The staging dotenv sets AUTH_USE_AUTH=true without a usable JWT secret,
and src/config.py validates the pair at import time — the same reason
unified-tests overrides it. This suite never runs the API server.
DEV-2035
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* test(live_llm): fix gpt-5.4 reasoning_effort and gemini replay-turn flake
- test_live_openai: gpt-5.4 dropped 'minimal' from the reasoning_effort
vocabulary, so the gpt5 caching test 400'd — and the OpenAI backend's
BadRequestError terminal swallowed it into an empty CompletionResult.
Pick the effort per model generation.
- test_live_tools_structured_output: use tool_choice='auto' on the
replay turn, matching the production dialectic loop (which never
forces 'none') — NONE mode is what provoked gemini-2.5-flash's empty
candidates. Drop the temperature pin so retries actually resample,
and treat a repeat tool call as a retryable attempt.
Verified live: full suite green, gemini 4/4 consecutive passes.
DEV-2035
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* ci: fail live LLM run when no staging secret was loaded
If the latest-tag fetch fails and no second tag exists, the fallback
step is skipped rather than failed, and the job would proceed without
provider keys — every test then skips via require_provider_key and the
run goes green. Guard on both fetch outcomes so that path fails loudly.
DEV-2035
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* docs(live-llm-tests-GHA): remove extra comments
* feat(structured-output): enable non-recursive schema references
* docs(structured-outputs): clean up new doc
* test(structured-output): fix caching refs memory leak, add tests
---------
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
|
||
|---|---|---|
| .. | ||
| README.md | ||
| __init__.py | ||
| conftest.py | ||
| model_matrix.py | ||
| test_live_anthropic.py | ||
| test_live_gemini.py | ||
| test_live_openai.py | ||
| test_live_structured_output_unions.py | ||
| test_live_tools_structured_output.py | ||
README.md
Live LLM Tests
These tests call real provider APIs and are disabled by default.
Run them with:
uv run pytest tests/live_llm -n 0 --live-llm --no-header -q
Required API key env vars:
LLM_ANTHROPIC_API_KEYLLM_OPENAI_API_KEYLLM_GEMINI_API_KEY
Model-family env vars:
LIVE_LLM_ANTHROPIC_45_PLUS_MODELSLIVE_LLM_OPENAI_GPT4_MODELSLIVE_LLM_OPENAI_GPT5_MODELSLIVE_LLM_OPENAI_OPENROUTER_NON_REASONING_MODELS(OpenAI-transport → OpenRouter-served non-reasoning models)LIVE_LLM_GEMINI_25_MODELSLIVE_LLM_GEMINI_30_MODELSLIVE_LLM_GEMINI_31_MODELS
Each model env var accepts a comma-separated list of bare model ids or provider-qualified ids.
Examples:
export LIVE_LLM_ANTHROPIC_45_PLUS_MODELS="claude-sonnet-4-5,claude-sonnet-4-6"
export LIVE_LLM_OPENAI_GPT4_MODELS="gpt-4.1"
export LIVE_LLM_OPENAI_GPT5_MODELS="gpt-5,gpt-5.4,gpt-5.4-mini"
export LIVE_LLM_OPENAI_OPENROUTER_NON_REASONING_MODELS="inception/mercury-2"
export LIVE_LLM_GEMINI_25_MODELS="gemini-2.5-flash,gemini-2.5-pro"
export LIVE_LLM_GEMINI_30_MODELS="gemini-3-flash-preview"
export LIVE_LLM_GEMINI_31_MODELS="gemini-3.1-pro-preview"
OpenRouter-routed models require additional env for the proxy endpoint:
export OPENROUTER_API_KEY="sk-or-v1-..."
# Per-feature config example:
# DERIVER_MODEL_CONFIG__TRANSPORT=openai
# DERIVER_MODEL_CONFIG__MODEL=inception/mercury-2
# DERIVER_MODEL_CONFIG__OVERRIDES__BASE_URL=https://openrouter.ai/api/v1
# DERIVER_MODEL_CONFIG__OVERRIDES__API_KEY_ENV=OPENROUTER_API_KEY
Coverage by provider:
- Anthropic: structured output path, prompt caching metrics, thinking blocks, multi-turn tool replay
- OpenAI GPT-4 class: structured outputs, prompt caching
- OpenAI GPT-5 class (incl. gpt-5.x point-releases): structured outputs, prompt caching,
reasoning_effort,max_completion_tokensrouting - OpenAI transport → OpenRouter non-reasoning models (e.g.
inception/mercury-2): non-chat / diffusion architectures must stay onmax_tokens, noreasoning_effort, tool-calling parameter-schema compatibility is the canary for exotic OR-served providers - Gemini 2.5/3.0 classes: structured outputs, cached-content reuse, thought signatures, multi-turn tool replay
- Gemini 3.1 class: thinking and tool replay coverage by default; structured-output/caching coverage should only be added once Google documents support for that path