"""Behavior tests for hermes_cli.inventory. Locks the invariants the three migrated consumers (web_server.py /api/model/options, tui_gateway model.options, tui_gateway model.save_key) depend on: - load_picker_context() reproduces the inline 17-LOC config-slice exactly. - with_overrides() is truthy-only (empty agent attrs must not clobber). - build_models_payload() returns a stable {providers, model, provider} shape and delegates curation to list_authenticated_providers (does not call provider_model_ids per row). - canonical_order keys on slug membership, not is_user_defined — section 3 of list_authenticated_providers sets is_user_defined=True for canonical slugs in the providers: dict, and that flag must NOT demote them to the tail. - picker_hints adds authenticated/auth_type/key_env/warning per row, matching the TUI ModelPickerDialog shape. """ from __future__ import annotations from unittest.mock import patch from hermes_cli.inventory import ( ConfigContext, build_models_payload, load_picker_context, ) # ─── load_picker_context ─────────────────────────────────────────────── def _cfg(model=None, providers=None, custom_providers=None) -> dict: return { "model": model if model is not None else {}, "providers": providers if providers is not None else {}, "custom_providers": custom_providers if custom_providers is not None else [], } def test_load_picker_context_coerces_numeric_yaml_provider(): """YAML parses unquoted `provider: 2070` as int; picker context must be str. Desktop GET /api/model/options crashed when a custom endpoint was named after a GPU: current_provider.strip() and providers dict keys .lower(). """ cfg = _cfg( model={ "provider": 2070, "default": "Qwen3.5-9B-Q4_K_M.gguf", "base_url": "http://192.168.1.10:8082/v1", }, providers={ 2070: { "name": 2070, "base_url": "http://192.168.1.10:8082/v1", "model": "Qwen3.5-9B-Q4_K_M.gguf", } }, ) with patch("hermes_cli.config.load_config", return_value=cfg): ctx = load_picker_context() assert ctx.current_provider == "2070" assert isinstance(ctx.current_provider, str) assert list(ctx.user_providers) == ["2070"] assert all(isinstance(k, str) for k in ctx.user_providers) assert ctx.current_model == "Qwen3.5-9B-Q4_K_M.gguf" assert ctx.current_base_url == "http://192.168.1.10:8082/v1" # ─── with_overrides ──────────────────────────────────────────────────── def _empty_ctx(provider="orig", model="orig-model", base_url="orig-url"): return ConfigContext( current_provider=provider, current_model=model, current_base_url=base_url, user_providers={}, custom_providers=[], ) # ─── build_models_payload ────────────────────────────────────────────── def _list_auth_returning(rows: list[dict]): """Patch list_authenticated_providers to return a fixed row list.""" return patch( "hermes_cli.model_switch.list_authenticated_providers", return_value=rows, ) def _nous_row(model: str = "openai/gpt-5.5") -> dict: return { "slug": "nous", "name": "Nous", "models": [model], "total_models": 1, "is_current": True, "is_user_defined": False, "source": "built-in", } def test_build_models_payload_returns_expected_shape(): rows = [ {"slug": "openrouter", "name": "OpenRouter", "models": ["m1"], "total_models": 1, "is_current": True, "is_user_defined": False, "source": "built-in"}, ] ctx = _empty_ctx(provider="openrouter", model="m1", base_url="") raw_config = { "moa": { "presets": { "default": {"enabled": True}, }, }, } with ( _list_auth_returning(rows), patch("hermes_cli.config.read_raw_config", return_value=raw_config), patch("hermes_cli.config.load_config", return_value=raw_config), ): payload = build_models_payload(ctx) assert set(payload.keys()) == {"providers", "model", "provider"} assert payload["model"] == "m1" assert payload["provider"] == "openrouter" assert payload["providers"][0]["slug"] == "moa" assert payload["providers"][0]["models"] == ["default"] assert payload["providers"][1:] == rows def test_build_models_payload_hides_moa_without_raw_preset(): """Strict opt-in (#63353): no ``moa`` row when the raw config has no enabled preset — the DEFAULT_CONFIG ``default`` preset everyone gets via ``load_config()`` is not a user choice.""" rows = [ {"slug": "openrouter", "name": "OpenRouter", "models": ["m1"], "total_models": 1, "is_current": True, "is_user_defined": False, "source": "built-in"}, ] ctx = _empty_ctx(provider="openrouter", model="m1", base_url="") # load_config() still returns the synthesized DEFAULT_CONFIG moa preset. synthesized = {"moa": {"presets": {"default": {"enabled": True}}}} with ( _list_auth_returning(rows), patch("hermes_cli.config.read_raw_config", return_value={}), patch("hermes_cli.config.load_config", return_value=synthesized), ): payload = build_models_payload(ctx) assert [r["slug"] for r in payload["providers"]] == ["openrouter"] def test_include_unconfigured_appends_canonical_skeletons(): """include_unconfigured=True adds CANONICAL_PROVIDERS rows that list_authenticated_providers didn't emit. Skeleton rows have empty models and source='canonical'. MoA is a virtual routing mode, not a canonical provider, so it is excluded — see issue #63353.""" rows = [ {"slug": "openrouter", "name": "OpenRouter", "models": ["m1"], "total_models": 1, "is_current": True, "is_user_defined": False, "source": "built-in"}, ] ctx = _empty_ctx(provider="openrouter") with _list_auth_returning(rows): payload = build_models_payload(ctx, include_unconfigured=True) # All canonical providers other than openrouter should appear as # skeleton rows. MoA is virtual/opt-in and excluded from unconfigured. from hermes_cli.models import CANONICAL_PROVIDERS seen_slugs = {r["slug"] for r in payload["providers"]} for entry in CANONICAL_PROVIDERS: if entry.slug == "moa": continue # virtual; only shown when explicitly configured assert entry.slug in seen_slugs, f"missing {entry.slug}" # Skeletons have empty models and source='canonical'. skeletons = [r for r in payload["providers"] if r.get("source") == "canonical"] assert all(r["models"] == [] for r in skeletons) assert all(r["total_models"] == 0 for r in skeletons) def test_explicit_only_filters_ambient_credentials_but_keeps_current_and_custom_rows(): rows = [ {"slug": "openai-codex", "name": "OpenAI Codex", "models": ["gpt-5.4"], "total_models": 1, "is_current": True, "is_user_defined": False, "source": "hermes"}, {"slug": "gemini", "name": "Gemini", "models": ["gemini-2.5-pro"], "total_models": 1, "is_current": False, "is_user_defined": False, "source": "built-in"}, {"slug": "copilot", "name": "Copilot", "models": ["gpt-5.4"], "total_models": 1, "is_current": False, "is_user_defined": False, "source": "hermes"}, {"slug": "nous", "name": "Nous", "models": ["anthropic/claude-sonnet-5"], "total_models": 1, "is_current": False, "is_user_defined": False, "source": "hermes"}, {"slug": "custom:lab", "name": "Lab", "models": ["lab-1"], "total_models": 1, "is_current": False, "is_user_defined": True, "source": "user-config"}, {"slug": "moa", "name": "MoA", "models": ["default"], "total_models": 1, "is_current": False, "is_user_defined": False, "source": "virtual"}, ] ctx = _empty_ctx(provider="openai-codex", model="gpt-5.4") with ( _list_auth_returning(rows), patch("hermes_cli.config.read_raw_config", return_value={}), patch( "hermes_cli.auth.is_provider_explicitly_configured", side_effect=lambda slug: slug == "gemini", ), ): payload = build_models_payload(ctx, explicit_only=True) assert [row["slug"] for row in payload["providers"]] == [ "openai-codex", "gemini", "custom:lab", ] def test_explicit_only_keeps_anthropic_row_with_oauth_credentials(): """Anthropic OAuth logins are deliberate sign-ins, not ambient credentials. Claude Code (~/.claude/.credentials.json) and Hermes' own device flow leave no trace in active_provider / model.provider / API-key env vars, so is_provider_explicitly_configured() returns False even though list_authenticated_providers just accepted those same credentials when building the row. The desktop explicit-only filter must keep it. """ rows = [ {"slug": "anthropic", "name": "Anthropic", "models": ["claude-sonnet-5"], "total_models": 1, "is_current": False, "is_user_defined": False, "source": "hermes"}, {"slug": "copilot", "name": "Copilot", "models": ["gpt-5.4"], "total_models": 1, "is_current": False, "is_user_defined": False, "source": "hermes"}, ] ctx = _empty_ctx(provider="opencode-go", model="glm-5.3") with ( _list_auth_returning(rows), patch("hermes_cli.config.read_raw_config", return_value={}), patch( "hermes_cli.auth.is_provider_explicitly_configured", return_value=False, ), patch( "hermes_cli.inventory._anthropic_oauth_credentials_present", return_value=True, ), ): payload = build_models_payload(ctx, explicit_only=True) slugs = [row["slug"] for row in payload["providers"]] assert "anthropic" in slugs, ( "Anthropic OAuth login must survive the explicit-only filter" ) assert "copilot" not in slugs, ( "ambient credential discovery must stay filtered" ) def test_explicit_only_drops_anthropic_row_without_oauth_credentials(): """No OAuth token and no explicit config -> Anthropic stays hidden.""" rows = [ {"slug": "anthropic", "name": "Anthropic", "models": ["claude-sonnet-5"], "total_models": 1, "is_current": False, "is_user_defined": False, "source": "hermes"}, ] ctx = _empty_ctx(provider="opencode-go", model="glm-5.3") with ( _list_auth_returning(rows), patch("hermes_cli.config.read_raw_config", return_value={}), patch( "hermes_cli.auth.is_provider_explicitly_configured", return_value=False, ), patch( "hermes_cli.inventory._anthropic_oauth_credentials_present", return_value=False, ), ): payload = build_models_payload(ctx, explicit_only=True) assert "anthropic" not in [row["slug"] for row in payload["providers"]] def test_anthropic_oauth_presence_accepts_pool_only_oauth_entry(): """A pool-only OAuth entry (auth.json credential_pool.anthropic) counts. Wired/device-flow tokens land in the credential pool, not in .anthropic_oauth.json or ~/.claude/.credentials.json. The presence check must accept them or the row is built and then silently dropped. """ from hermes_cli.inventory import _anthropic_oauth_credentials_present with ( patch( "agent.anthropic_credentials.read_hermes_oauth_credentials", return_value=None, ), patch( "agent.anthropic_credentials.read_claude_code_credentials", return_value=None, ), patch( "hermes_cli.auth.read_credential_pool", return_value=[ {"auth_type": "oauth", "access_token": "sk-ant-oat01-pool"} ], ), ): assert _anthropic_oauth_credentials_present() is True # api_key pool entries are NOT OAuth logins — presence must stay False # (they are handled by the explicit-config gate / env var paths). with ( patch( "agent.anthropic_credentials.read_hermes_oauth_credentials", return_value=None, ), patch( "agent.anthropic_credentials.read_claude_code_credentials", return_value=None, ), patch( "hermes_cli.auth.read_credential_pool", return_value=[ {"auth_type": "api_key", "access_token": "sk-ant-api03-key"} ], ), ): assert _anthropic_oauth_credentials_present() is False # ─── picker_hints ────────────────────────────────────────────────────── def test_picker_hints_marks_authed_rows_authenticated(): rows = [ {"slug": "openrouter", "name": "OpenRouter", "models": ["m1"], "total_models": 1, "is_current": True, "is_user_defined": False, "source": "built-in"}, ] ctx = _empty_ctx() with _list_auth_returning(rows): payload = build_models_payload(ctx, picker_hints=True) assert payload["providers"][0]["authenticated"] is True def test_picker_hints_api_key_warning_format(): """For api_key providers with a defined env var, the warning must point to that env var.""" rows = [] ctx = _empty_ctx() with _list_auth_returning(rows): payload = build_models_payload( ctx, include_unconfigured=True, picker_hints=True, ) # anthropic uses api_key + ANTHROPIC_API_KEY. anthropic = next( r for r in payload["providers"] if r["slug"] == "anthropic" ) assert "ANTHROPIC_API_KEY" in anthropic["warning"] # ─── canonical_order ─────────────────────────────────────────────────── def test_canonical_order_uses_slug_not_is_user_defined_flag(): """Section 3 of list_authenticated_providers sets is_user_defined=True for canonical slugs that appear in the providers: config dict. canonical_order MUST key on slug membership, not the flag — otherwise canonical providers configured via the keyed schema get demoted to the tail. """ from hermes_cli.models import CANONICAL_PROVIDERS canonical_slug = CANONICAL_PROVIDERS[2].slug # any canonical rows = [ # A truly-custom row (correct: is_user_defined=True) {"slug": "custom:Ollama", "name": "Ollama", "models": [], "total_models": 0, "is_current": False, "is_user_defined": True, "source": "user-config"}, # A canonical row that the substrate flagged as user-defined # because the user configured it via providers: dict. {"slug": canonical_slug, "name": "x", "models": ["m1"], "total_models": 1, "is_current": False, "is_user_defined": True, "source": "built-in"}, ] ctx = _empty_ctx() with _list_auth_returning(rows): payload = build_models_payload(ctx, canonical_order=True) slugs = [r["slug"] for r in payload["providers"]] # Canonical-slug row must come BEFORE truly-custom rows, regardless # of is_user_defined. canonical_idx = slugs.index(canonical_slug) custom_idx = slugs.index("custom:Ollama") assert canonical_idx < custom_idx, ( f"canonical {canonical_slug} demoted to tail " f"(canonical_idx={canonical_idx} > custom_idx={custom_idx})" ) def test_canonical_order_with_unconfigured_preserves_full_universe(): """Combined picker call: include_unconfigured + picker_hints + canonical_order is the production TUI shape. Verify the result has CANONICAL_PROVIDERS in declaration order, hints applied, custom rows trailing. MoA is excluded from unconfigured skeletons (virtual, opt-in only — issue #63353). """ from hermes_cli.models import CANONICAL_PROVIDERS rows = [ {"slug": "custom:Ollama", "name": "Ollama", "models": [], "total_models": 0, "is_current": False, "is_user_defined": True, "source": "user-config"}, ] ctx = _empty_ctx() with _list_auth_returning(rows): payload = build_models_payload( ctx, include_unconfigured=True, picker_hints=True, canonical_order=True, ) slugs = [r["slug"] for r in payload["providers"]] # First row: first canonical provider in declaration order. assert slugs[0] == CANONICAL_PROVIDERS[0].slug # Custom row trails all visible canonical rows. MoA is virtual/opt-in # so it is excluded from the unconfigured skeleton set. visible_canonical_count = sum( 1 for e in CANONICAL_PROVIDERS if e.slug != "moa" ) assert slugs.index("custom:Ollama") >= visible_canonical_count # ─── Integration: end-to-end through real load_picker_context ────────── def test_end_to_end_with_real_context_no_credentials_leak(monkeypatch): """Full pipeline: real load_picker_context + real list_authenticated_providers. Verify no credential string ever appears in the returned payload, even with picker_hints=True.""" canary = "sk-canary-XYZ-must-not-appear" monkeypatch.setenv("OPENROUTER_API_KEY", canary) monkeypatch.setenv("ANTHROPIC_API_KEY", canary) cfg = _cfg(model={"provider": "openrouter"}) with patch("hermes_cli.config.load_config", return_value=cfg): ctx = load_picker_context() payload = build_models_payload( ctx, include_unconfigured=True, picker_hints=True, ) import json as _json assert canary not in _json.dumps(payload) def test_payload_shape_compatible_with_modelpickerdialog_frontend(): """Frontend (web/src/components/ModelPickerDialog.tsx) reads: name, slug, models, total_models, is_current, warning, authenticated. Verify every authenticated/skeleton row exposes those keys. """ rows = [ {"slug": "openrouter", "name": "OpenRouter", "models": ["m1"], "total_models": 1, "is_current": True, "is_user_defined": False, "source": "built-in"}, ] ctx = _empty_ctx() with _list_auth_returning(rows): payload = build_models_payload( ctx, include_unconfigured=True, picker_hints=True, ) required_keys = {"name", "slug", "models", "total_models", "is_current", "authenticated"} for row in payload["providers"]: missing = required_keys - row.keys() assert not missing, f"row {row['slug']} missing keys: {missing}" # ─── Aggregator dedup (issue #45954) ─────────────────────────────────── def _user_provider_row(slug: str, models: list[str]) -> dict: return { "slug": slug, "name": slug.title(), "models": models, "total_models": len(models), "is_current": False, "is_user_defined": True, "source": "user-config", } def _aggregator_row(slug: str, models: list[str]) -> dict: return { "slug": slug, "name": slug.title(), "models": models, "total_models": len(models), "is_current": False, "is_user_defined": False, "source": "built-in", } def test_user_defined_rows_carry_alias_set_for_gui_current_match(): """Custom provider rows must expose `aliases` so the desktop picker can match a session's canonical `custom:` identity against the row's bare-key slug (#87035). Built-in rows carry no aliases. """ rows = [ { "slug": "myep", "name": "My Endpoint", "models": ["my-model"], "total_models": 1, "is_current": True, "is_user_defined": True, "source": "user-config", "api_url": "http://localhost:8000/v1", }, _nous_row() | {"is_current": False}, ] ctx = _empty_ctx(provider="custom:myep", model="my-model") with _list_auth_returning(rows): payload = build_models_payload(ctx) by_slug = {r["slug"]: r for r in payload["providers"]} aliases = by_slug["myep"]["aliases"] # The canonical session identity must be matchable via the alias set. assert "custom:myep" in aliases assert "myep" in aliases assert "custom:my-endpoint" in aliases assert "aliases" not in by_slug["nous"] def test_aggregator_dedup_removes_overlapping_models(): """Models served by a user-defined provider are removed from aggregator rows so the picker doesn't show them under the wrong provider. (#45954)""" rows = [ _user_provider_row("litellm-proxy", [ "nvidia/nim/minimax-m3", "nvidia/nim/kimi-k2.6", ]), _aggregator_row("openrouter", [ "minimax/minimax-m3", "nvidia/nim/minimax-m3", # overlaps with litellm-proxy "anthropic/claude-sonnet-4.6", ]), ] ctx = _empty_ctx() with _list_auth_returning(rows): payload = build_models_payload(ctx) or_row = next(r for r in payload["providers"] if r["slug"] == "openrouter") proxy_row = next(r for r in payload["providers"] if r["slug"] == "litellm-proxy") # User-defined provider keeps all its models assert proxy_row["models"] == ["nvidia/nim/minimax-m3", "nvidia/nim/kimi-k2.6"] # Aggregator lost the overlapping model but kept the rest assert "nvidia/nim/minimax-m3" not in or_row["models"] assert "minimax/minimax-m3" in or_row["models"] assert "anthropic/claude-sonnet-4.6" in or_row["models"] assert or_row["total_models"] == 2 def test_flat_namespace_reseller_keeps_first_party_models_overlapping_user_proxy(): """opencode-go / opencode-zen are flagged ``is_aggregator=True`` (their flat ``/v1/models`` returns bare IDs the model-switch resolver searches), but they are NOT routing aggregators — every model they list is a first-party model under the user's subscription. When a user also runs a custom proxy that happens to serve a same-named model, the picker dedup must NOT strip the reseller's own catalog. Regression for #47077, where opencode-go showed only 13 of 19 models because minimax-m3/m2.7/m2.5, glm-5/5.1, and deepseek-v4-flash were deduped against an overlapping custom provider. """ rows = [ _user_provider_row("custom:my-proxy", [ "minimax-m3", "minimax-m2.7", "glm-5", "deepseek-v4-flash", ]), _aggregator_row("opencode-go", [ "kimi-k2.6", "minimax-m3", "minimax-m2.7", "glm-5", "deepseek-v4-flash", "qwen3.7-max", ]), _aggregator_row("openrouter", ["minimax-m3", "anthropic/claude-sonnet-4.6"]), ] ctx = _empty_ctx() with _list_auth_returning(rows): payload = build_models_payload(ctx) go_row = next(r for r in payload["providers"] if r["slug"] == "opencode-go") or_row = next(r for r in payload["providers"] if r["slug"] == "openrouter") # The reseller keeps ALL of its first-party models — nothing stripped. assert go_row["models"] == [ "kimi-k2.6", "minimax-m3", "minimax-m2.7", "glm-5", "deepseek-v4-flash", "qwen3.7-max", ] assert go_row["total_models"] == 6 # A TRUE routing aggregator is still deduped against the user's models. assert "minimax-m3" not in or_row["models"] assert "anthropic/claude-sonnet-4.6" in or_row["models"] def test_build_models_payload_no_max_models_returns_full_list(): """When max_models is not passed (None), build_models_payload must return the full model list — not truncate to the old default of 50. Regression for #48279: Kilo Gateway picker was capped at 50 of 336 models, making most models undiscoverable via search.""" full_models = [f"model-{i}" for i in range(100)] rows = [ { "slug": "kilocode", "name": "Kilo Code", "models": full_models, "total_models": len(full_models), "is_current": False, "is_user_defined": False, "source": "built-in", }, ] ctx = _empty_ctx() with _list_auth_returning(rows): # No max_models argument — should return all 100 models payload = build_models_payload(ctx) kilo_row = next(r for r in payload["providers"] if r["slug"] == "kilocode") assert kilo_row["models"] == full_models assert kilo_row["total_models"] == 100 assert len(kilo_row["models"]) == 100 # ─── refresh flag (cache-bust) ───────────────────────────────────────── def test_list_authenticated_providers_refresh_busts_cache(): """refresh=True clears the provider-model disk cache exactly once; refresh=False leaves it untouched (so normal picker opens stay snappy).""" from hermes_cli import model_switch with patch("hermes_cli.models.clear_provider_models_cache") as clear: model_switch.list_authenticated_providers(refresh=False) assert clear.call_count == 0 model_switch.list_authenticated_providers(refresh=True) assert clear.call_count == 1 def test_picker_metadata_uses_one_config_read_for_real_models_dev_lookups(tmp_path, monkeypatch): """Custom-provider metadata stays constant-read as its model count grows (#119048). ``get_model_capabilities`` and ``get_model_info`` deliberately stay real: each lookup resolves ``providers.lab.catalog_provider`` before consulting the seeded models.dev catalog. Removing snapshot threading from that path makes the larger payload re-open config.yaml once per lookup. """ from agent import models_dev from hermes_cli import config as config_module home = tmp_path / "hermes" home.mkdir() (home / "config.yaml").write_text( "providers:\n" " lab:\n" " catalog_provider: openrouter\n" "model_overrides: {}\n", encoding="utf-8", ) monkeypatch.setenv("HERMES_HOME", str(home)) config_module._LOAD_CONFIG_CACHE.clear() config_module._RAW_CONFIG_CACHE.clear() models = [ "openai/model-a", "anthropic/model-b", "openai/model-c", "anthropic/model-d", "openai/model-e", "anthropic/model-f", "openai/model-g", "anthropic/model-h", ] registry = { "openrouter": { "models": { model: { "id": model, "tool_call": True, "reasoning": model.startswith("openai/"), "release_date": f"2026-01-{index:02d}", "limit": {"context": 200000, "output": 8192}, } for index, model in enumerate(models, start=1) }, }, } monkeypatch.setattr(models_dev, "_models_dev_cache", registry) monkeypatch.setattr(models_dev, "_models_dev_cache_time", float("inf")) snapshot = config_module.load_config_readonly() baseline = [ ( models_dev.get_model_capabilities("custom:lab", model), models_dev.get_model_info("custom:lab", model), ) for model in models ] snapshot_result = [ ( models_dev.get_model_capabilities("custom:lab", model, config=snapshot), models_dev.get_model_info("custom:lab", model, config=snapshot), ) for model in models ] assert snapshot_result == baseline def _rows(model_ids): return [{ "slug": "custom:lab", "name": "Lab", "models": model_ids, "total_models": len(model_ids), "is_current": False, "is_user_defined": True, "source": "user-config", }] def _build_and_count(model_ids): cfg_get_calls = 0 real_cfg_get = models_dev._cfg_get def counted_cfg_get(*keys, **kwargs): nonlocal cfg_get_calls cfg_get_calls += 1 return real_cfg_get(*keys, **kwargs) with ( _list_auth_returning(_rows(model_ids)), patch("hermes_cli.inventory._local_runtime_row", return_value=None), patch("hermes_cli.inventory._moa_provider_row", return_value=None), patch("hermes_cli.models.model_supports_fast_mode", return_value=False), patch("hermes_cli.inventory._reasoning_catalog_reader", return_value=None), patch.object(models_dev, "_cfg_get", side_effect=counted_cfg_get), patch.object( config_module, "load_config_readonly", wraps=config_module.load_config_readonly, ) as readonly_load, patch.object( config_module, "_load_config_impl", wraps=config_module._load_config_impl, ) as config_impl, ): payload = build_models_payload( _empty_ctx(), capabilities=True, featured=True, ) return payload, cfg_get_calls, readonly_load.call_count, config_impl.call_count small_payload, small_cfg_get, small_reads, small_impls = _build_and_count(models[:3]) large_payload, large_cfg_get, large_reads, large_impls = _build_and_count(models) # Snapshot threading leaves _cfg_get's cheap dict traversals proportional # to model count, while eliminating config/signature reads from the hot path. assert small_cfg_get < large_cfg_get assert (small_reads, large_reads) == (1, 1) assert (small_impls, large_impls) == (1, 1) small_row = small_payload["providers"][0] large_row = large_payload["providers"][0] assert small_row["capabilities"] == { model: {"fast": False, "reasoning": model.startswith("openai/")} for model in models[:3] } assert large_row["featured_models"] == models