shinnmen / internal /runtime /executor /aistudio_executor.go

Commit History

  1. refactor(executor): split providers and preserve legacy entrypoints⁠

    shinmentakezo07 +1: Claude Opus 4.6
    3e5ab39

  1. feat: passthrough upstream response headers to clients⁠

    Kirill Turanskiy
    409fbf7

  1. refactor(translator, executor): remove redundant `bytes.Clone` calls for improved performance⁠

    Unverified signature
    1b0fbba
  2. refactor(executor): optimize payload cloning and streamline SDK translator usage⁠

    Unverified signature
    c702df8

  1. feat(openai): add responses/compact support

    Shady Khalifa
    Unverified signature
    8f7449a

  1. feat(executor): apply payload rules using requested model

    hkfires
    69a5340

  1. feat(registry): support provider-specific model info lookup

    hkfires
    4771da8

  1. refactor(thinking): pass source and target formats to ApplyThinking for cross-format validation⁠

    hkfires
    4072514

  1. fix(executor): properly handle thinking application errors

    hkfires
    e934726
  2. refactor: improve thinking logic

    hkfires
    ee9e572

  1. feat(executor): add HttpRequest support across executors for better http request handling

    Unverified signature
    a64baa3

  1. feat(executor): enhance payload translation with original request context⁠

    Unverified signature
    ae057e3

  1. refactor(executor): remove redundant upstream model parameter from translateRequest

    hkfires
    8628ab7
  2. refactor(executor): resolve upstream model at conductor level before execution

    hkfires
    c37b7ba
  3. fix(executor): use upstream model for thinking config and payload translation

    hkfires
    7559da4

  1. fix(gemini): add optional skip for gemini3 thinking conversion

    hkfires
    baafbd7

  1. fix: apply thinkingLevel from model suffix metadata for Gemini 3⁠

    Ben Vargas
    a1a3463
  2. feat: use thinkingLevel for Gemini 3 models per Google documentation⁠

    Ben Vargas
    965592d

  1. refactor(thinking): export thinking helpers⁠

    hkfires
    ca524b4

  1. Fixed: #534⁠

    hkfires
    4744928

  1. refactor(executor): clarify executor comments and oauth names

    hkfires
    9cc01d9
  2. refactor(executor): clarify providers and streams⁠

    hkfires
    f3f3cbf

  1. fix(executor): centralize default thinking config

    hkfires
    c60202a
  2. feat(executor): normalize thinking budget across all Gemini executors

    hkfires
    ab09327
  3. feat(aistudio): normalize thinking budget in request translation

    hkfires
    f8fe943

  1. refactor(executor): dedupe thinking metadata helpers across Gemini executors⁠

    auroraflux
    82379d5

  1. fix(aistudio): strip Gemini generation config overrides⁠

    hkfires
    8f128c6

  1. Fixed: #302⁠

    Unverified signature
    edb9a12

  1. feat(executor): share SSE usage filtering across streams

    hkfires
    7dbb8b8

  1. Fixed: #291⁠

    Unverified signature
    028d53f

  1. **feat(runtime): add payload configuration support for executors**⁠

    Unverified signature
    a917f9b

  1. fix(executor): remove `safetySettings` from payload in token counting request

    Unverified signature
    3ffa428

  1. fix(executor): only apply thinking config to supported models

    hkfires
    c896fd9

  1. fix(executor): pass authID to relay instead of provider

    hkfires
    c3b5398
  2. feat(registry): unify Gemini models and add AI Studio set

    hkfires
    f56bc9a

  1. fix(aistudio): ensure colon-spaced JSON in responses

    hkfires
    7af3f21
  2. refactor(wsrelay): rename RoundTrip to NonStream

    hkfires
    965993c
  3. fix(aistudio): remove generationConfig and tools when action is countTokens

    hkfires
    ce3727e

  1. fix(aistudio): strip usage metadata from non-final stream chunks

    hkfires
    c140413
  2. feat(aistudio): track Gemini usage and improve stream errors

    hkfires
    8c00322
  3. feat: add websocket routing and executor unregister API⁠

    hkfires
    7ed60dd