mirror of
https://github.com/danny-avila/LibreChat.git
synced 2026-11-12 09:45:47 +00:00
b4afbb3a1e
* 🛝 perf: Render Chats Independently of Catalog Discovery
* Preserve Target Authorization and Passive Startup Intent
* Keep Startup Selection, Readiness, and Catalog Keys Consistent
* Preserve Remembered Targets and Validate Scoped Presets
* Retain Remembered Assistant Resource Aliases Across Endpoint Switches
* Apply URL Targets Before Waiting for Send Readiness
* Fix First-Page Agent Discovery and Await Model Selection
* Preserve Spec Access and Refresh Selected Assistant State
* Preserve Builder Deletion State and Start History Reads Earlier
* fix: Preserve Startup Recovery and History Read Ownership
* fix: Retain Express Query Routing in Assistant Lookups
* fix: Prioritize Route Reads and Recover Credential Changes
* fix: Bind Credential Revocation to Its Requested Scope
* fix: Refresh Credential Metadata After Pending Key Reads
* fix: Revalidate Inactive Credential Metadata on Target Mount
* fix: Preserve Reviewed Chat Resolutions After Rebase
* fix: Resolve Dev Conflicts and Preserve Startup Boundaries
* test: Preserve the ChatView Slot Target Boundary
* fix: Defer Terminal Delivery Until Subscription Attachment Completes
* fix: Contain Deferred Terminal Callback Failures
* test: Classify Historical Tool Name Capability Reads
* fix: Activate Requested Catalogs and Isolate Provider Failures
* fix: Keep Healthy Provider Choices Available During Catalog Recovery
* test: Exclude Copied Trace Bundles From Dependency Comparison
---------
Co-authored-by: Lia <lia@librechat.ai>
10 lines
363 B
JavaScript
10 lines
363 B
JavaScript
const { createTokenConfigController } = require('@librechat/api');
|
|
const { getRequestModelsConfig } = require('~/server/controllers/ModelController');
|
|
const { getValueKey, getMultiplier, getCacheMultiplier } = require('~/models');
|
|
|
|
module.exports = createTokenConfigController({
|
|
getRequestModelsConfig,
|
|
getValueKey,
|
|
getMultiplier,
|
|
getCacheMultiplier,
|
|
});
|