can1357 fb3534740f feat(catalog): added GLM-5.2 reasoning support for ZAI and zhipu completions ...

- Added ZAI GLM-5.2 reasoning-effort mapping, translating minimal to none and xhigh to max.
- Enabled ZAI and zhipu GLM-5.2 completion requests to send reasoning_effort and tool_stream.
- Added provider token clamping so GLM-5.2 completion requests use capped max_tokens.
- Updated catalog policies to route GLM-5.2 max-token and reasoning support through ZAI/zhipu hosts.
- Removed synthetic HF model entries and aligned GLM-5.2 catalog specs with real providers.

Fixes #2833

2026-06-17 09:38:43 +02:00

..

azure-provider.test.ts

feat(catalog/provider-models): added OpenAI model provider descriptors

2026-06-15 10:40:48 +02:00

build.test.ts

feat(catalog): stripped gateway prefixes and promo tags from model display names

2026-06-12 07:37:38 +02:00

canonical-limit-fallback.test.ts

fix(coding-agent): avoided blocking auth resolution during model switches

2026-06-13 15:39:04 +02:00

descriptors.test.ts

feat(catalog): updated provider defaults and model priorities for new versions

2026-06-14 08:43:40 +02:00

discovery-null-limits.test.ts

fix: handled unknown model limits as null to avoid artificial token caps

2026-06-13 15:35:40 +02:00

fireworks-serverless-discovery.test.ts

feat(cross-cutting): added multi-syntax in-band tool-call support for runtime tool conversion

2026-06-15 07:33:24 +02:00

generated-policies.test.ts

fix(catalog): pinned minimax m3 context

2026-06-14 16:45:23 +00:00

github-copilot-model-limits.test.ts

feat(catalog): added requestModelId routing and Copilot long-context discovery support

2026-06-11 21:28:32 +02:00

github-copilot-wire.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

google-vertex-discovery.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

hosts.test.ts

feat(cross-cutting): centralized model compatibility handling with catalog resolvers

2026-06-10 04:39:42 +02:00

identity-family.test.ts

fix(catalog): classify GLM in modelFamilyToken family tokens

2026-06-14 17:31:59 +02:00

issue-772-repro.test.ts

Drop Xiaomi ASR models from catalog

2026-06-15 05:35:19 -07:00

issue-830-repro.test.ts

fix(catalog): moved OpenAI effort maps into thinking metadata

2026-06-11 23:28:07 +00:00

issue-847-repro.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

issue-887-repro.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

issue-1617-repro.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

issue-1846-repro.test.ts

feat: added build-time compatibility in ModelSpec/buildModel pipeline

2026-06-10 06:20:51 +02:00

issue-1849-repro.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

issue-2105-repro.test.ts

feat(catalog/provider-models): added OpenAI model provider descriptors

2026-06-15 10:40:48 +02:00

issue-2113-repro.test.ts

fix(catalog): corrected moonshot kimi request shaping

2026-06-11 05:28:17 +00:00

issue-2299-repro.test.ts

fix(catalog): routed nvidia-hosted qwen to chat_template_kwargs thinking

2026-06-11 10:04:09 +00:00

issue-2558-repro.test.ts

test(catalog): sourced Model and ModelSpec types from pi-catalog

2026-06-14 11:14:18 +00:00

minimax-bundled-catalog.test.ts

fix(catalog): pinned minimax m3 context

2026-06-14 16:45:23 +00:00

model-id-affixes.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

model-provider-priority.test.ts

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

model-thinking.test.ts

feat(catalog): added GLM-5.2 reasoning support for ZAI and zhipu completions

2026-06-17 09:38:43 +02:00

nanogpt-model-limits.test.ts

refactor(catalog): baked thinking metadata into buildModel pipeline

2026-06-10 07:22:11 +02:00

ollama-cloud-output-caps.test.ts

refactor(catalog): baked thinking metadata into buildModel pipeline

2026-06-10 07:22:11 +02:00

ollama-cloud-provider.test.ts

feat(catalog/provider-models): added OpenAI model provider descriptors

2026-06-15 10:40:48 +02:00

ollama-provider.test.ts

fix(catalog): moved OpenAI effort maps into thinking metadata

2026-06-11 23:28:07 +00:00

preferred-dialect.test.ts

refactor: renamed grammar module to dialect with unified transcript rendering

2026-06-15 13:58:12 +02:00

umans-provider.test.ts

fix umans gateway websearch handling

2026-06-16 01:14:43 -07:00

variant-collapse.test.ts

feat(catalog): added thinking.requiresEffort baking for mandatory-reasoning upstreams

2026-06-12 08:20:27 +02:00

venice-provider.test.ts

fix umans max token cap

2026-06-15 23:38:48 -07:00

wafer.test.ts

feat: added build-time compatibility in ModelSpec/buildModel pipeline

2026-06-10 06:20:51 +02:00

xai-oauth-bundle.test.ts

fix: handled unknown model limits as null to avoid artificial token caps

2026-06-13 15:35:40 +02:00

zai-bundled-catalog.test.ts

test(catalog): fix zai catalog test types

2026-06-13 13:04:23 -07:00

zenmux-provider.test.ts

test(catalog): update zenmux default model expectation

2026-06-14 17:27:26 +02:00

zhipu-compat.test.ts

feat(catalog): added GLM-5.2 reasoning support for ZAI and zhipu completions

2026-06-17 09:38:43 +02:00