mirror of
https://github.com/MoonshotAI/kimi-code.git
synced 2026-11-08 17:28:58 +00:00
* refactor(agent-core-v2): clarify human-layer names * refactor(agent-core-v2): name provider-holding members after their role - LlmRequestConfig/LlmRecoveryContext/adapter Model: credentials -> credentialProvider - requester/provider convertError option -> classifyError (matches LlmErrorClassifier) - kimiOAuthCredentialProvider -> createKimiOAuthCredentialProvider - ModelCatalog.buildCredentials -> buildCredentialProvider * refactor(agent-core-v2): rename the format-layer error hook to classifyErrorHook threads the requester's classifyError option through under a matching name * refactor(agent-core-v2): finish the human-layer naming pass - turn.failure.classified -> turn.failure.recovery.evaluated - recovery proposal beforeRetry -> beforeNextAttempt - llmRequester currentCredentials/credentialsForTurn -> currentCredentialProvider/credentialProviderForTurn - unexport ProtocolTraitsByName (only backs ProtocolTraitFor) * refactor(agent-core-v2): shorten the turn failure event to turn.failure.evaluated |
||
|---|---|---|
| .. | ||
| basic.ts | ||
| context-usage.ts | ||
| identity.ts | ||
| kimi-select-tools.ts | ||
| kosong-config-stress.ts | ||
| model-requester-boundary.ts | ||
| smoke.ts | ||