Andreas/LibreChat
1
0
Fork 0
mirror of https://github.com/danny-avila/LibreChat.git synced 2025-09-22 06:00:56 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 3

🔧 refactor: Improve Agent Context & Minor Fixes (#5349)

Browse source 
* refactor: Improve Context for Agents

* 🔧 fix: Safeguard against undefined properties in OpenAIClient response handling

* refactor: log error before re-throwing for original stack trace

* refactor: remove toolResource state from useFileHandling, allow svg files

* refactor: prevent verbose logs from axios errors when using actions

* refactor: add silent method recordTokenUsage in AgentClient

* refactor: streamline token count assignment in BaseClient

* refactor: enhance safety settings handling for Gemini 2.0 model

* fix: capabilities structure in MCPConnection

* refactor: simplify civic integrity threshold handling in GoogleClient and llm

* refactor: update token count retrieval method in BaseClient tests

* ci: fix test for svg
This commit is contained in:
Danny Avila 2025-01-17 12:55:48 -05:00 • committed by GitHub
parent e309c6abef
commit b35a8b78e2
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
19 changed files with 324 additions and 112 deletions
Download patch file Download diff file Expand all files Collapse all files

26
api/app/clients/BaseClient.js
View file

@ -4,6 +4,7 @@ const {
supportsBalanceCheck,
isAgentsEndpoint,
isParamEndpoint,
EModelEndpoint,
ErrorTypes,
Constants,
CacheKeys,
@ -11,6 +12,7 @@ const {
} = require('librechat-data-provider');
const { getMessages, saveMessage, updateMessage, saveConvo } = require('~/models');
const { addSpaceIfNeeded, isEnabled } = require('~/server/utils');
const { truncateToolCallOutputs } = require('./prompts');
const checkBalance = require('~/models/checkBalance');
const { getFiles } = require('~/models/File');
const { getLogStores } = require('~/cache');
@ -95,7 +97,7 @@ class BaseClient {
* @returns {number}
*/
getTokenCountForResponse(responseMessage) {
logger.debug('`[BaseClient] recordTokenUsage` not implemented.', responseMessage);
logger.debug('[BaseClient] `recordTokenUsage` not implemented.', responseMessage);
}
/**
@ -106,7 +108,7 @@ class BaseClient {
* @returns {Promise<void>}
*/
async recordTokenUsage({ promptTokens, completionTokens }) {
logger.debug('`[BaseClient] recordTokenUsage` not implemented.', {
logger.debug('[BaseClient] `recordTokenUsage` not implemented.', {
promptTokens,
completionTokens,
});
@ -287,6 +289,9 @@ class BaseClient {
}
async handleTokenCountMap(tokenCountMap) {
if (this.clientName === EModelEndpoint.agents) {
return;
}
if (this.currentMessages.length === 0) {
return;
}
@ -394,6 +399,21 @@ class BaseClient {
_instructions && logger.debug('[BaseClient] instructions tokenCount: ' + tokenCount);
let payload = this.addInstructions(formattedMessages, _instructions);
let orderedWithInstructions = this.addInstructions(orderedMessages, instructions);
if (this.clientName === EModelEndpoint.agents) {
const { dbMessages, editedIndices } = truncateToolCallOutputs(
orderedWithInstructions,
this.maxContextTokens,
this.getTokenCountForMessage.bind(this),
);
if (editedIndices.length > 0) {
logger.debug('[BaseClient] Truncated tool call outputs:', editedIndices);
for (const index of editedIndices) {
payload[index].content = dbMessages[index].content;
}
orderedWithInstructions = dbMessages;
}
}
let { context, remainingContextTokens, messagesToRefine, summaryIndex } =
await this.getMessagesWithinTokenLimit(orderedWithInstructions);
@ -625,7 +645,7 @@ class BaseClient {
await this.updateUserMessageTokenCount({ usage, tokenCountMap, userMessage, opts });
} else {
responseMessage.tokenCount = this.getTokenCountForResponse(responseMessage);
completionTokens = this.getTokenCount(completion);
completionTokens = responseMessage.tokenCount;
}
await this.recordTokenUsage({ promptTokens, completionTokens, usage });