AnythingLLM
Anything LLM Pack.#
This Integration is part of theSupported versions
Supported Cortex XSOAR versions: 6.10.0 and later.
Retrieval Augmented Generation (RAG) with LLM and Vector DB that can be local for full data privacy or cloud-based for greater functionality
#
Configure AnythingLLM in CortexParameter | Description | Required |
---|---|---|
AnythingLLM URL (e.g., http://<url to local AnythingLLM>:3001) or https://<Anything LLM cloud url> | True | |
AnythingLLM API Key | True | |
Cloudflare Access Client Id | False | |
Cloudflare Access Client Secret | False | |
AIRS API URL | Leave blank if AIRS is not being used | False |
AIRS API Key | Leave blank if AIRS is not being used | False |
Sync Scan LLM Prompts with AIRS | False | |
Sync Scan LLM Response with AIRS | False | |
AIRS Profile Name for Scan Requests | The profile must be created in the AIRS console | False |
AIRS Application Name for Scan Requests | False | |
AIR User for Scan Requests | False | |
Anything LLM Model in use for Scan Requests | False |
#
CommandsYou can execute these commands from the CLI, as part of an automation, or in a playbook. After you successfully execute a command, a DBot message appears in the War Room with the command details.
#
anyllm-document-upload-fileUploads an XSOAR file entry to the custom-documents folder.
#
Base Commandanyllm-document-upload-file
#
InputArgument Name | Description | Required |
---|---|---|
fileentry | XSOAR file entry to upload - example: 181@24789. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-document-upload-linkUploads a web link to the custom-documents folder
#
Base Commandanyllm-document-upload-link
#
InputArgument Name | Description | Required |
---|---|---|
link | Web link to upload - example: https://unit42.paloaltonetworks.com/darkgate-malware-uses-excel-files". | Required |
title | No description provided. | Required |
description | No description provided. | Required |
author | No description provided. | Required |
source | No description provided. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-document-upload-textUpload text content as a document to the custom-documents folder
#
Base Commandanyllm-document-upload-text
#
InputArgument Name | Description | Required |
---|---|---|
text | Raw text content that is the document. | Required |
title | Document title to use when uploading. | Required |
description | Description of the document. | Optional |
author | Author of the document. | Optional |
source | Source of the document. | Optional |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-newCreates a new workspace in AnythingLLM
#
Base Commandanyllm-workspace-new
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace to create. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-deleteDeletes an AnythingLLM workspace
#
Base Commandanyllm-workspace-delete
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace to delete. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-listList all the workspaces in AnythingLLM
#
Base Commandanyllm-workspace-list
#
InputArgument Name | Description | Required |
---|
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-getGet a specific workspace details
#
Base Commandanyllm-workspace-get
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-settingsUpdate workspace settings
#
Base Commandanyllm-workspace-settings
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Required |
settings | JSON object for the settings. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-add-embeddingAdd a document to a workspace and create its vector embedding in the workspace
#
Base Commandanyllm-workspace-add-embedding
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Required |
folder | Folder name containing the document. | Required |
document | Document name to add as an embedding. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-delete-embeddingDelete a document embedding from the workspace
#
Base Commandanyllm-workspace-delete-embedding
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Required |
folder | Folder the document originated from. | Required |
document | Name of the document to have its embedding deleted. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-document-createfolderCreate a new document folder
#
Base Commandanyllm-document-createfolder
#
InputArgument Name | Description | Required |
---|---|---|
folder | Name of the folder to create. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-document-moveMove a document from a source folder to a destination folder
#
Base Commandanyllm-document-move
#
InputArgument Name | Description | Required |
---|---|---|
srcfolder | Name of the source folder. | Required |
dstfolder | Name of the destination folder. | Optional |
document | Document name to move. | Optional |
#
Context OutputThere is no context output for this command.
#
anyllm-document-deleteDelete a document
#
Base Commandanyllm-document-delete
#
InputArgument Name | Description | Required |
---|---|---|
folder | Name of the folder. | Required |
document | Name of the document to delete. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-chatSend a chat message to a workspace (default thread). Query mode is based on embedded documents in chat, whereas chat mode is more general.
#
Base Commandanyllm-workspace-chat
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Required |
message | Message to send. | Required |
mode | Mode to chat, query or chat. Possible values are: query, chat. | Required |
format | No description provided. Possible values are: markdown, dictionary. Default is dictionary. | Optional |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-stream-chatSend a stream chat message to a workspace (default thread). Query mode is based on embedded documents in chat, whereas chat mode is more general
#
Base Commandanyllm-workspace-stream-chat
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Required |
message | Message to send. | Required |
mode | Chat mode, query or chat. Possible values are: query, chat. | Optional |
#
Context OutputThere is no context output for this command.
#
anyllm-document-listList all document details
#
Base Commandanyllm-document-list
#
InputArgument Name | Description | Required |
---|
#
Context OutputThere is no context output for this command.
#
anyllm-document-getGet a specific document details
#
Base Commandanyllm-document-get
#
InputArgument Name | Description | Required |
---|---|---|
folder | Folder containing the document. | Required |
document | Document name. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-pinSet the pinned status of a document embedding
#
Base Commandanyllm-workspace-pin
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Workspace name. | Required |
folder | Folder the document originated from. | Required |
document | Document name. | Required |
status | Set pin status to true or false. Possible values are: true, false. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-thread-chatsGet the conversation for a workspace thread
#
Base Commandanyllm-workspace-thread-chats
#
InputArgument Name | Description | Required |
---|---|---|
workspace | No description provided. | Required |
thread | No description provided. | Optional |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-thread-chatSend a chat a message to a conversation thread
#
Base Commandanyllm-workspace-thread-chat
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Required |
thread | Name of the conversation thread. | Required |
message | Message to send. | Required |
mode | Mode to chat, query or chat. Possible values are: query, chat. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-thread-newCreate a new conversation thread
#
Base Commandanyllm-workspace-thread-new
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Required |
thread | Name of the new conversation thread. | Required |
#
Context OutputThere is no context output for this command.
#
anyllm-workspace-thread-deleteDelete a thread in a workspace
#
Base Commandanyllm-workspace-thread-delete
#
InputArgument Name | Description | Required |
---|---|---|
workspace | Name of the workspace. | Optional |
thread | Name of the thread. | Optional |
#
Context OutputThere is no context output for this command.
#
anyllm-scan-requestPerform an AI security scan of a request with AIRS
#
Base Commandanyllm-scan-request
#
InputArgument Name | Description | Required |
---|---|---|
profile_name | AIRS profile name to use in scanning. | Required |
app_name | AIRS application name - metadata. | Optional |
app_user | AIRS application user name - metadata. | Optional |
ai_model | LLM model in use - metadata. | Optional |
user_ip | User IP - metadata. | Optional |
prompt | LLM prompt to scan. | Optional |
response | LLM response to scan. | Optional |
code_prompt | LLM prompt with code to scan. | Optional |
code_response | LLM response with code to scan. | Optional |
context | LLM context for grounding. | Optional |
#
Context OutputThe context data returned from a scan request is:
Key | Value |
---|---|
action | block or allow |
category | malicious etc |
profile_id | ID of the AIRS profile used |
profile_name | Name of the AIRS profile used |
prompt_detected | Details of a prompt detection - dictionary |
response_detected | Details of a response detection - dictionary |
scan_id | Scan ID |
tr_id | TR ID |