https://docs.infinilabs.com/coco-server/v0.11.0/Recent content in Coco Server on Coco Server | The AI-powered Enterprise Search, all in one spaceHugo -- gohugo.iohttps://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/file_metadata/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/file_metadata/File Metadata Processor # Extracts metadata from supported file types and stores the results in the document metadata. File type Extracted metadata Image colors (top-3 dominant color names), width (px), height (px) Configuration # Parameter Type Required Default Description message_field string No messages Pipeline context key for the input messages output_queue object No null Queue to push processed documents to Example # - file_metadata: output_queue: name: "documents_with_metadata"https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/attachment_processor/generate_cover/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/attachment_processor/generate_cover/Generate Attachment Cover Processor # Generates a cover image and a thumbnail for an attachment. The generated cover and thumbnail are stored in the attachment metadata and can be retrieved via the attachment API. Supported format Notes PDF Cover from the first page PPTX / DOCX / XLSX and OpenDocument equivalents Cover from the first page Markdown Rendered cover Image Thumbnail of the image itself Requirements/Dependencies # Tool Required for pdftoppm (poppler-utils) PDF cover generation LibreOffice (soffice) Office document cover generation Chromium (headless) Markdown rendered cover Configuration # Parameter Type Required Default Description message_field string No messages Pipeline context key for the input messages Example # - generate_attachment_cover: {}https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/attachment_processor/text_extraction/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/attachment_processor/text_extraction/Attachment Text Extraction Processor # Extracts text content from an attachment and stores the result as the attachment’s searchable text. For image attachments, the processor uses a vision model to generate a text description. For document attachments, it uses Apache Tika text extraction. This processor does not extract embedded attachments from the attachment being processed. Embedded images inside PDFs, PPTX files, and similar formats are removed from the extracted text instead of being uploaded as child attachments.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/generate_cover/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/generate_cover/Generate Document Cover Processor # Generates a cover image and a thumbnail for a document. Note: For attachment cover generation, use generate_attachment_cover instead. The generated cover and thumbnail are stored with the document and can be retrieved via the document API. Supported format Notes PDF Cover from the first page PPTX / DOCX / XLSX and OpenDocument equivalents Cover from the first page Markdown Rendered cover Image Thumbnail of the image itself Requirements/Dependencies # Tool Required for pdftoppm (poppler-utils) PDF cover generation LibreOffice (soffice) Office document cover generation Chromium (headless) Markdown rendered cover Configuration # Parameter Type Required Default Description message_field string No messages Pipeline context key for the input messages output_queue object No null Queue to push processed documents to Example # - generate_document_cover: output_queue: name: "documents_with_cover"https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/document_text_attachment_extraction/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/document_text_attachment_extraction/Document Text Attachment Extraction Processor # Extracts the full text content from a document and splits it into chunks. When extract_attachments is enabled, embedded images are extracted and linked as attachments. When it is disabled, the processor extracts text only. Note: For attachment text extraction, use attachment_text_extraction instead. Requirements # Service Required for Apache Tika server Text extraction from PDF, DOCX, XLSX, and other Tika-backed formats; also used for embedded-image OCR when extract_attachments is enabled Supported formats # Supported format Notes PDF Text and OCR for image-only pages PPTX / PPT / PPTM Per-slide text and embedded image OCR Image Description generated by a vision model DOCX, XLSX and other office formats Plain-text extraction Configuration # Parameter Type Required Default Description message_field string No messages Pipeline context key for the input messages output_queue object No null Queue to push processed documents to tika_endpoint string No http://127.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/face_extraction/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/face_extraction/Face Extraction Processor # Detects human faces in a document and attempts to identify each person using a vision model. Detected faces and their identities are stored in the document metadata. Requirements/Dependencies # Dependency Required for Pigo facefinder cascade binary Face detection (set via pigo_facefinder_path) Apache Tika server Extracting embedded images from PDF, PPTX, DOCX Supported formats # Supported format Notes Image Faces detected directly PDF, PPTX, DOCX Faces detected in embedded images Depends on document_text_attachment_extraction — this processor must run after document_text_attachment_extraction so that the extracted text is available as context for the vision model.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/configuration/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/configuration/Configuration Reference # Coco Server is configured through a YAML file (coco.yml). This reference documents all available configuration options with detailed examples. Environment Variables # Environment variables can be defined in the env section and referenced elsewhere in the configuration using $[[env.VAR_NAME]]. env: ES_ENDPOINT: https://localhost:9200 ES_USERNAME: admin ES_PASSWORD: $[[keystore.ES_PASSWORD]] WEB_BINDING: 0.0.0.0:9000 API_BINDING: 0.0.0.0:2900 Sensitive values can be stored securely using the keystore: ./bin/coco keystore add ES_PASSWORD Coco Server Settings # Core server settings are defined under the coco section.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/document_summarization/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/document_summarization/Document Summarization Processor # Generates an AI summary and structured insights for a document using a language model. Documents shorter than min_input_document_length or longer than max_input_document_length are skipped. Requirements # A configured language model provider is required. Set model_provider and model in the processor config, or configure a default language model in the application settings. Configuration # Parameter Type Required Default Description message_field string No messages Pipeline context key for the input messages output_queue object No null Queue to push processed documents to model_provider string No (app default) Language model provider ID model string No (app default) Language model name model_context_length int No — Model context window size in tokens (minimum 4000) min_input_document_length int No 100 Skip documents shorter than this (bytes) max_input_document_length int No 100000 Skip documents longer than this (bytes) ai_insights_max_length int No 500 Target length for the generated summary (tokens) llm_generation_lang string No (app default) BCP 47 language tag for generated content (e.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/document_embedding/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/document_embedding/Document Embedding Processor # Generates vector embeddings for a document’s text chunks using an embedding model to enable semantic search and retrieval. Requirements # A configured embedding model provider is required. Set model_provider and model in the processor config, or configure a default embedding model in the application settings. The processor always produces vectors with a fixed dimension of 1024, which is the only dimension supported by Coco semantic search.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/extract_tags/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/extract_tags/Extract Tags Processor # Extracts structured keyword tags from a document’s content using a language model. Documents that have not yet been summarized are skipped. Run document_summarization before this processor. Requirements # A configured language model provider is required. Set model_provider and model in the processor config, or configure a default language model in the application settings. Configuration # Parameter Type Required Default Description message_field string No messages Pipeline context key for the input messages output_queue object No null Queue to push processed documents to model_provider string No (app default) Language model provider ID model string No (app default) Language model name model_context_length int No — Model context window size in tokens (minimum 4000) llm_generation_lang string No (app default) BCP 47 language tag for generated content (e.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/authentication/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/authentication/Authentication # Authentication Methods # The API supports three methods of authentication: 1. Login API # Authenticate with the Coco Server to obtain a JWT access token. Password-only login (single-user mode): curl -H 'Content-Type: application/json' -XPOST http://localhost:9000/account/login -d'{ "password": "mypassword" }' Email and password login (multi-user mode): curl -H 'Content-Type: application/json' -XPOST http://localhost:9000/account/login -d'{ "email": "admin@example.com", "password": "mypassword" }' Response: { "access_token": "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9...", "expire_in": 86400, "id": "coco-default-user", "status": "ok", "username": "coco-default-user" } Field Type Description access_token string JWT token for Bearer authentication (valid for 24 hours).https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/google_drive/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/google_drive/Google Drive Connector # Register Google Drive Connector # curl -XPOST "http://localhost:9000/connector" -d '{ "name": "Google Drive Connector", "description": "Fetch the files metadata from Google Drive.", "icon": "/assets/connector/google_drive/icon.png", "path_hierarchy":false, "category": "cloud_storage", "tags": [ "google", "storage" ], "url": "http://coco.rs/connectors/google_drive", "assets": { "icons": { "default": "/assets/connector/google_drive/icon.png", "audio": "/assets/connector/google_drive/audio.png", "form": "/assets/connector/google_drive/form.png", "document": "/assets/connector/google_drive/document.png", "drawing": "/assets/connector/google_drive/drawing.png", "folder": "/assets/connector/google_drive/folder.png", "fusiontable": "/assets/connector/google_drive/fusiontable.png", "jam": "/assets/connector/google_drive/jam.png", "map": "/assets/connector/google_drive/map.png", "ms_excel": "/assets/connector/google_drive/ms_excel.png", "ms_powerpoint": "/assets/connector/google_drive/ms_powerpoint.png", "ms_word": "/assets/connector/google_drive/ms_word.png", "pdf": "/assets/connector/google_drive/pdf.png", "photo": "/assets/connector/google_drive/photo.https://docs.infinilabs.com/coco-server/v0.11.0/docs/getting-started/install/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/getting-started/install/Coco Server User Guide # Run Coco Server with Docker Manual install Coco Server 1. Quick Start (Recommended for Most Users) # This method is the simplest way to get Coco Server running. It uses Docker-managed volumes, which handles data persistence automatically without requiring manual directory setup on your host machine. Command: docker run -d \ --name cocoserver \ -p 9000:9000 \ -v coco_data_vol:/app/easysearch/data \ -v coco_config_vol:/app/easysearch/config \ -v coco_logs_vol:/app/easysearch/logs \ infinilabs/coco:0.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/access_token/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/access_token/API Token # API Tokens are long-lived tokens that can be used in your own applications to access Coco Server APIs. Tokens are valid for 365 days from creation. Use the X-API-TOKEN header to authenticate requests. API Token Fields # Field Type Description name string Custom name for the token. Auto-generated if not provided. permissions array[object] List of permission keys the token is authorized for.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/hugo_site/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/hugo_site/Hugo Site Connector # Register Hugo Site Connector # curl -XPOST "http://localhost:9000/connector" -d '{ "name": "Hugo Site Connector", "description": "Fetch the index.json file from a specified Hugo site.", "icon": "/assets/connector/hugo_site/icon.png", "category": "website", "tags": [ "static_site", "hugo", "web" ], "url": "http://coco.rs/connectors/hugo_site", "assets": { "icons": { "default": "/assets/connector/hugo_site/web.png", "blog": "/assets/connector/hugo_site/blog.png", "web": "/assets/connector/hugo_site/web.png", "web_page": "/assets/connector/hugo_site/web_page.png", "news": "/assets/connector/hugo_site/news.png" } }, "processor":{ "enabled":true, "name":"hugo_site" } }' Make sure hugo generated the json format with this:https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/rss/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/rss/Rss Connector # Register RSS Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name" : "RSS Connector", "description" : "Fetch items from a specified RSS feed.", "category" : "website", "icon" : "/assets/icons/connector/rss/icon.png", "tags" : [ "rss", "feed", "web" ], "url" : "http://coco.rs/connectors/rss", "assets" : { "icons" : { "default" : "/assets/icons/connector/rss/icon.png" } }, "processor":{ "enabled":true, "name":"rss" } }' Use the RSS Connector # //request curl -H 'Content-Type: application/json' -XPOST "http://localhost:9000/datasource/" -d' { "name":"My RSS feed", "type":"connector", "connector":{ "id":"rss's connector id", "config":{ "urls": [ "The RSS link" ] } } }' Below is the config parameters supported by this connector.https://docs.infinilabs.com/coco-server/v0.11.0/docs/getting-started/setup/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/getting-started/setup/Setup Wizard # Open your browser and enter http://localhost:9000/#/guide to access the initialization wizard, as shown below: Enter your username, email, and password, then click Next, as shown: Select the LLM type: DeepSeek, Ollama, or OpenAI Configure the LLM endpoint Set the default model Enable keepalive and set an appropriate interval Provide the token Click Next to complete the initialization. Login # After initialization, you’ll be redirected to the login page, as shown:https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/confluence/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/confluence/Confluence Connector # Register Confluence Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "Confluence wiki Connector", "description": "Fetch Confluence Wiki pages and blogposts.", "category": "website", "icon": "/assets/icons/connector/confluence/icon.png", "tags": [ "wiki", "storage", "docs", "web" ], "url": "http://coco.rs/connectors/confluence", "assets": { "icons": { "default": "/assets/icons/connector/confluence/icon.png" } }, "processor": { "enabled": true, "name": "confluence" } }' Use the Confluence Connector # The Confluence Connector allows you to index data from your Confluence wiki into your system.https://docs.infinilabs.com/coco-server/v0.11.0/docs/tutorials/howto_create_your_own_datasource/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/tutorials/howto_create_your_own_datasource/How to integrate Coco AI with your own datasource # This guide will walk you through the process of creating a custom datasource and building a simple connector to integrate it with the system. Obtain API Token # Make sure you have a valid API token before continue. Create a simple connector # curl -H'X-API-TOKEN: cv9bb63q50k5eikkq43gp07lmg60gn2nor7uf650xpxazd6a6nns8uttymz52uvnplhmtaf829x71lph51li' -XPOST http://localhost:9000/connector/ -d'{ "name": "My Connector" }' Response { "_id": "cv9bc23q50k5eikkq4c0", "result": "created" } Create a datasource for this connector # curl -H'X-API-TOKEN: cv9bb63q50k5eikkq43gp07lmg60gn2nor7uf650xpxazd6a6nns8uttymz52uvnplhmtaf829x71lph51li' -H 'Content-Type: application/json' -XPOST http://localhost:9000/datasource/ -d' { "name":"My Datasource", "type":"connector", "connector":{ "id":"cv9bc23q50k5eikkq4c0" } }' Responsehttps://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/local_fs/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/local_fs/Local FS Connector # Register Local FS Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name" : "Local Filesystem Connector", "description" : "Scan and fetch metadata from local files.", "category" : "local_storage", "icon" : "/assets/icons/connector/local_fs/icon.png", "tags" : [ "storage", "filesystem" ], "url" : "http://coco.rs/connectors/local_fs", "assets" : { "icons" : { "default" : "/assets/icons/connector/local_fs/icon.png" } } }' Use the Local FS Connector # The Local FS Connector allows you to index data from your local filesystem into your system.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/network_drive/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/network_drive/Network Drive Connector # Register Network Drive Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "Network Drive Connector", "description": "Scan and extract metadata from network shared files.", "category": "cloud_storage", "icon": "/assets/icons/connector/network_drive/icon.png", "tags": [ "filesystem", "storage", "web" ], "url": "http://coco.rs/connectors/network_drive", "assets": { "icons": { "default": "/assets/icons/connector/network_drive/icon.png" } }, "processor": { "enabled": true, "name": "network_drive" } }' Use the Network Drive Connector # The Network Drive Connector allows you to index data from your SMB/CIFS shares into your system.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/notion/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/notion/Notion Connector # Register Notion Connector # curl -XPOST "http://localhost:9000/connector" -d '{ "name": "Notion Docs Connector", "description": "Fetch the docs metadata for notion.", "icon": "/assets/connector/notion/icon.png", "category": "website", "tags": [ "docs", "notion", "web" ], "url": "http://coco.rs/connectors/notion", "assets": { "icons": { "default": "/assets/connector/notion/icon.png", "web_page": "/assets/connector/notion/icon.png", "database": "/assets/connector/notion/database.png", "page": "/assets/connector/notion/page.png" } }, "processor":{ "enabled":true, "name":"notion" } }' Use the Notion Connector # The Notion Connector allows you to index data from your notion account into your system.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/s3/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/s3/S3 Connector # Register S3 Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "S3 Object Storage Connector", "description": "Fetch file metadata from S3 cloud storage.", "category": "cloud_storage", "icon": "/assets/icons/connector/s3/icon.png", "tags": [ "s3", "storage" ], "url": "http://coco.rs/connectors/s3", "assets": { "icons": { "default": "/assets/icons/connector/s3/icon.png" } }, "processor": { "enabled": true, "name": "s3" } }' Use the S3 Connector # The S3 Connector allows you to index data from your S3 service into your system.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/yuque/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/yuque/Yuque Connector # Register Yuque Connector # curl -XPOST "http://localhost:9000/connector" -d '{ "name": "Yuque Docs Connector", "description": "Fetch the docs metadata for yuque.", "icon": "/assets/connector/yuque/icon.png", "category": "website", "tags": [ "static_site", "hugo", "web" ], "url": "http://coco.rs/connectors/hugo_site", "assets": { "icons": { "default": "/assets/connector/yuque/icon.png", "book": "/assets/connector/yuque/book.png", "board": "/assets/connector/yuque/board.png", "sheet": "/assets/connector/yuque/sheet.png", "table": "/assets/connector/yuque/table.png", "doc": "/assets/connector/yuque/doc.png" } }, "processor":{ "enabled":true, "name":"yuque" } }' Use the Yuque Connector # The Yuque Connector allows you to index data from your Yuque account into your system.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/jira/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/jira/Jira Connector # Register Jira Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "Jira Connector", "description": "Fetch Jira issues, comments, and attachments from your Jira projects.", "category": "project_management", "icon": "/assets/icons/connector/jira/icon.png", "tags": [ "project", "issue_tracking", "agile", "collaboration" ], "url": "http://coco.rs/connectors/jira", "assets": { "icons": { "default": "/assets/icons/connector/jira/icon.png" } }, "processor": { "enabled": true, "name": "jira" } }' Use the Jira Connector # The Jira Connector allows you to index issues, comments, and attachments from your Jira projects into your system.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/feishu/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/feishu/Feishu/Lark Connector # The Feishu/Lark connector indexes cloud documents from Feishu and Lark, including documents, spreadsheets, mind notes, multi-dimensional tables, and knowledge bases. Features # 🔍 Smart Search: Keyword-based search for cloud documents 📚 Multiple Document Types: Support for doc, sheet, slides, mindnote, bitable, file, docx, folder, shortcut 🔐 Dual Authentication: OAuth 2.0 and user access token authentication (choose one) ⚡ Efficient Sync: Scheduled and manual synchronization 🔄 Recursive Search: Automatically search folder contents recursively 🔄 Token Auto-refresh: OAuth authentication supports automatic refresh of access_token and refresh_token 🌐 Dynamic Redirect: Supports dynamic OAuth redirect URI construction for multi-environment deployment 🏗️ Unified Architecture: Feishu and Lark share base implementation with 95% code reuse 📁 Directory Access: Hierarchical browsing based on Feishu’s original folder structure; folder directories are created on-the-fly 🕒 Incremental Sync: Robust last-modified tracking with safety buffer to avoid edge misses Supported Platforms # Feishu # Domain: open.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/assistant/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/assistant/Assistant # Assistant API Reference # Below is the field description for the assistant. Field Type Description name string The assistant’s name. description string A brief description of the assistant. icon string The icon representing the assistant in the UI. type string The assistant type, e.g., deep_think. config.intent_analysis_model object Model configuration for intent analysis.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/document/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/document/Document API # Document API Reference # Documents # Below is the field description for the document. Field Type Description source object The source of the document. category string Primary category of the document, e.g., report. subcategory string Secondary category of the document, e.g., 2024. categories array[string] List of categories the document belongs to, e.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/password/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/password/Modify Password # Modify the current user’s password. Requires authentication. Modify Password # //request curl -XPUT http://localhost:9000/account/password \ -H "Authorization: Bearer <access_token>" \ -H 'Content-Type: application/json' \ -d'{ "old_password":"current_password", "new_password":"new_secure_password" }' //response { "_id": "coco-default-user", "result": "updated" } Parameters # Field Type Required Description old_password string Yes The user’s current password. new_password string Yes The new password to set.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/mysql/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/mysql/MySQL Connector # Register MySQL Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "MySQL Connector", "description": "Fetch data from MySQL database.", "category": "database", "icon": "/assets/icons/connector/mysql/icon.png", "tags": [ "sql", "storage", "web" ], "url": "http://coco.rs/connectors/mysql", "assets": { "icons": { "default": "/assets/icons/connector/mysql/icon.png" } }, "processor": { "enabled": true, "name": "mysql" } }' Use the MySQL Connector # The MySQL Connector allows you to index data from your database by executing a custom SQL query.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/postgresql/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/postgresql/PostgreSQL Connector # Register PostgreSQL Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "PostgreSQL Connector", "description": "Fetch data from PostgreSQL database.", "category": "database", "icon": "/assets/icons/connector/postgresql/icon.png", "tags": [ "sql", "storage", "web" ], "url": "http://coco.rs/connectors/postgresql", "assets": { "icons": { "default": "/assets/icons/connector/postgresql/icon.png" } }, "processor": { "enabled": true, "name": "postgresql" } }' Use the PostgreSQL Connector # The PostgreSQL Connector allows you to index data from your database by executing a custom SQL query.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/profile/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/profile/Profile API # Profile API Reference # Profile # Below is the field description for the profile object. Field Type Description id string Unique identifier for the user profile. name string User’s display name. email string User’s email address. permissions array[string] List of permission keys assigned to the user, e.g., ["coco:document:read"]. Get Profile # Requires authentication.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/search/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/search/Search API # Search API Reference # Parameters # Parameter Type Default Description query string "" The search query string. from int 0 The starting offset of the search results (used for pagination). size int 10 The number of search results to return. datasource string "" Filter results by datasource ID. category string "" Filter results by primary category.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/mssql/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/mssql/MS SQL Connector # Register MS SQL Connector # curl -XPOST "http://localhost:9000/connector" -d ' { "name" : "Microsoft SQL Server Connector", "description" : "Fetch data from Microsoft SQL Serve database.", "category" : "database", "icon" : "/assets/icons/connector/mssql/icon.png", "tags" : [ "sql", "storage", "web" ], "url" : "http://coco.rs/connectors/mssql", "assets" : { "icons" : { "default" : "/assets/icons/connector/mssql/icon.png" } }, "processor": { "enabled": true, "name": "mssql" } }' Use the MS SQL Connector # The MS SQL Connector allows you to index data from your database by executing a custom SQL query.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/oracle/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/oracle/Oracle Connector # Register Oracle Connector # curl -XPOST "http://localhost:9000/connector/" -d ' { "name" : "Oracle Connector", "description" : "Fetch data from Oracle database.", "category" : "database", "icon" : "/assets/icons/connector/oracle/icon.png", "tags" : [ "sql", "storage", "web" ], "url" : "http://coco.rs/connectors/oracle", "assets" : { "icons" : { "default" : "/assets/icons/connector/oracle/icon.png" } }, "processor":{ "enabled":true, "name":"oracle" } }' Use the Oracle Connector # The Oracle Connector allows you to index data from your database by executing a custom SQL query.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/dropbox/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/dropbox/Dropbox Connector # Register Dropbox Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "Dropbox Connector", "description": "Index files and folders from Dropbox", "icon": "/assets/icons/connector/dropbox/icon.png", "category": "cloud_storage", "path_hierarchy": false, "tags": [ "dropbox", "cloud_storage", "file_sharing" ], "url": "http://coco.rs/connectors/dropbox", "assets": { "icons": { "default": "/assets/icons/connector/dropbox/icon.png", "docx": "/assets/icons/connector/dropbox/docx.png", "xlsx": "/assets/icons/connector/dropbox/xlsx.png", "pptx": "/assets/icons/connector/dropbox/pptx.png", "pdf": "/assets/icons/connector/dropbox/pdf.png", "photo": "/assets/icons/connector/dropbox/photo.png", "zip": "/assets/icons/connector/dropbox/zip.png", "document": "/assets/icons/connector/dropbox/document.png", "paper": "/assets/icons/connector/dropbox/paper.png", "gdoc": "/assets/icons/connector/dropbox/gdoc.png", "gexcel": "/assets/icons/connector/dropbox/gexcel.png", "gppt": "/assets/icons/connector/dropbox/gppt.png" } }, "processor": { "enabled": true, "name": "dropbox" } }' Use the Dropbox Connector # The Dropbox Connector allows you to index files and folders from Dropbox.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/github/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/github/GitHub Connector # Register GitHub Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "GitHub Connector", "description": "Fetch repositories, issues, and pull requests from GitHub.", "icon": "/assets/icons/connector/github/icon.png", "category": "website", "tags": [ "git", "code", "vcs", "website" ], "url": "http://coco.rs/connectors/github", "assets": { "icons": { "default": "/assets/icons/connector/github/icon.png", "repository": "/assets/icons/connector/github/repository.png", "issue": "/assets/icons/connector/github/issue.png", "pull_request": "/assets/icons/connector/github/pull_request.png" } }, "processor": { "enabled": true, "name": "github" } }' Use the GitHub Connector # The GitHub Connector allows you to index repositories, issues, and pull requests from your GitHub account or organization.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/logout/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/account/logout/Logout # Logout API # The Logout API securely logs the user out from the Coco Server. It destroys the current session. Both GET and POST methods are supported. //request curl -XPOST http://localhost:9000/account/logout \ -H "Authorization: Bearer <access_token>" //response { "status": "ok" }https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/neo4j/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/neo4j/Neo4j Connector # Register Neo4j Connector # curl -XPOST "http://localhost:9000/connector/" -d ' { "name" : "Neo4j Connector", "description" : "Fetch data from Neo4j graph database using Cypher.", "category" : "database", "icon" : "/assets/icons/connector/neo4j/icon.png", "tags" : [ "graph", "database", "neo4j" ], "url" : "http://coco.rs/connectors/neo4j", "assets" : { "icons" : { "default" : "/assets/icons/connector/neo4j/icon.png" } }, "processor": { "enabled": true, "name": "neo4j" } }' Use the Neo4j Connector # The Neo4j connector runs a Cypher query, transforms the resulting rows into documents, and streams them into coco-server.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/salesforce/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/salesforce/Salesforce Connector # Register Salesforce Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "Salesforce Connector", "description": "Fetch data from Salesforce with intelligent field caching and query optimization.", "icon": "/assets/icons/connector/salesforce/icon.png", "category": "crm", "tags": [ "salesforce", "crm", "api", "oauth" ], "url": "http://coco.rs/connectors/salesforce", "assets": { "icons": { "default": "/assets/icons/connector/salesforce/icon.png", "account": "/assets/icons/connector/salesforce/account.png", "opportunity": "/assets/icons/connector/salesforce/opportunity.png", "contact": "/assets/icons/connector/salesforce/contact.png", "lead": "/assets/icons/connector/salesforce/lead.png", "campaign": "/assets/icons/connector/salesforce/campaign.png", "case": "/assets/icons/connector/salesforce/case.png" } }, "processor": { "enabled": true, "name": "salesforce" } }' Use the Salesforce Connector # The Salesforce Connector allows you to index data from your Salesforce org with intelligent field caching, query optimization, and comprehensive data extraction.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/gitlab/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/gitlab/GitLab Connector # Register GitLab Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "GitLab Connector", "description": "Fetch repositories, issues, merge requests, wikis, and snippets from GitLab.", "icon": "/assets/icons/connector/gitlab/icon.png", "category": "website", "tags": [ "git", "code", "vcs", "website" ], "url": "http://coco.rs/connectors/gitlab", "assets": { "icons": { "default": "/assets/icons/connector/gitlab/icon.png", "repository": "/assets/icons/connector/gitlab/repository.png", "issue": "/assets/icons/connector/gitlab/issue.png", "merge_request": "/assets/icons/connector/gitlab/merge_request.png", "wiki": "/assets/icons/connector/gitlab/wiki.png", "snippet": "/assets/icons/connector/gitlab/snippet.png" } }, "processor": { "enabled": true, "name": "gitlab" } }' Use gitlab as a unique identifier, as it is a builtin connector.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/gitea/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/gitea/Gitea Connector # Register Gitea Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "Gitea Connector", "description": "Fetch repositories, issues, pull requests, and wikis from Gitea.", "icon": "/assets/icons/connector/gitea/icon.png", "category": "website", "tags": [ "git", "code", "vcs", "website" ], "url": "http://coco.rs/connectors/gitea", "assets": { "icons": { "default": "/assets/icons/connector/gitea/icon.png", "repository": "/assets/icons/connector/gitea/repository.png", "issue": "/assets/icons/connector/gitea/issue.png", "pull_request": "/assets/icons/connector/gitea/pull_request.png" } }, "processor": { "enabled": true, "name": "gitea" } }' Use the Gitea Connector # The Gitea Connector allows you to index repositories, issues, pull requests, and wikis from your Gitea instance.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/box/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/box/Box Cloud Storage Connector # Register Box Connector # curl -XPOST "http://localhost:9000/connector/" -d '{ "name": "Box Cloud Storage Connector", "description": "Index files and folders from Box, supporting both Free and Enterprise accounts with multi-user access.", "icon": "/assets/icons/connector/box/icon.png", "category": "cloud_storage", "path_hierarchy": false, "tags": [ "box", "cloud_storage", "file_sharing" ], "url": "http://coco.rs/connectors/box", "assets": { "icons": { "default": "/assets/icons/connector/box/icon.png", "bookmark": "/assets/icons/connector/box/bookmark.png", "boxcanvas": "/assets/icons/connector/box/boxcanvas.png", "boxnote": "/assets/icons/connector/box/boxnote.png", "docx": "/assets/icons/connector/box/docx.png", "excel-spreadsheet": "/assets/icons/connector/box/excel-spreadsheet.png", "google-docs": "/assets/icons/connector/box/google-docs.png", "google-sheets": "/assets/icons/connector/box/google-sheets.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/milvus/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/milvus/Milvus Connector # Register Milvus Connector # curl -XPOST "http://localhost:9000/connector/" -d ' { "name" : "Milvus Vector Database Connector", "description" : "Fetch vector and scalar data from Milvus collections with incremental sync and pagination.", "category" : "vector_database", "icon" : "/assets/icons/connector/milvus/icon.png", "tags" : [ "vector", "database", "milvus", "vector_store" ], "url" : "http://coco.rs/connectors/milvus", "assets" : { "icons" : { "default" : "/assets/icons/connector/milvus/icon.png" } }, "processor": { "enabled": true, "name": "milvus" } }' Use the Milvus Connector # The Milvus connector queries a Milvus collection, retrieves vectors and scalar fields, and streams them as documents into coco-server.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/mongodb/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connectors/mongodb/MongoDB Connector # Register MongoDB Connector # curl -XPUT "http://localhost:9000/connector/" -d ' { "name" : "MongoDB Connector", "description" : "Fetch data from MongoDB collections with incremental sync support.", "category" : "database", "icon" : "/assets/icons/connector/mongodb/icon.png", "tags" : [ "nosql", "database", "mongodb" ], "url" : "http://coco.rs/connectors/mongodb", "assets" : { "icons" : { "default" : "/assets/icons/connector/mongodb/icon.png" } }, "processor": { "enabled": true, "name": "mongodb" } }' Use mongodb as the unique identifier because it is a built-in connector.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connector/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/connector/Connector # Work with Connector # A Connector defines how to access and fetch data from external services such as Google Drive, GitHub, Confluence, databases, and more. Connectors are used by datasources to pull data into Coco Server. Connector API # Below is the field description for the connector. Field Type Description name string The connector’s name. description string A brief description of the connector.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/datasource/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/datasource/Datasource # Work with Datasource # Datasource defines where the data comes from, usually we can use a specify connector to fetch data from a specify datasource. Below is the field description for the datasource. Field Type Description name string The datasource’s name. type string The datasource type, e.g., connector. description string A brief description of the datasource.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/integration/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/integration/Integration # Work with Integration # The integration generates a piece of code based on certain configuration parameters, which can be embedded into other websites. This code allows you to quickly use Coco AI’s search and chat capabilities. Integration API # Below is the field description for the integration. Field Type Description name string The integration’s name. type string The integration type.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/model_provider/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/model_provider/LLM Provider # Work with LLM Provider # The LLM (Large Language Model) Provider enables seamless integration of various AI models into your application. It supports multiple model types, including Deepseek, OpenAI, and more. This guide provides a comprehensive overview of how to effectively utilize the LLM Provider. LLM Provider API # Below is the field description for the model provider. Field Type Description name string The model provider’s name.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/mcp_server/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/mcp_server/MCP Server # Work with MCP Server # The MCP Server provides commonly used MCP configurations for large model calls. MCP Server API # Below is the field description for the MCP Server. Field Type Description name string The MCP Server’s name. type string The type to access the API of the MCP Server, possible values: stdio, streamable http.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/attachment/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/attachment/Attachment # Attachment API are used to upload your local files to Coco server. Upload attachment # curl -X POST http://localhost:9000/attachment/_upload -F "files=LOCAL_FILE" Example curl -X POST http://localhost:9000/attachment/_upload \ -H "X-API-TOKEN: cv9pnurq50k1hii28630jy429g4b49viecrlj9529onpa6n0lti7yohioitvyotd0677rop5uszc0cnll03j" \ -F "files=@/Users/medcl/Downloads/tmp/neurips19-diskann.pdf" \ -F "files=@/Users/medcl/Downloads/tmp/Adaptive_searching_in_succinctly_encoded.pdf" The session_id need to be replaced with actual session id. Response { "acknowledged": true, "attachments": [ "cv9q94bq50k2r0s6nob0", "cv9q94bq50k2r0s6nobg" ] } Download attachment # curl -X GET http://localhost:9000/attachment/cv9q94bq50k2r0s6nobg \ -H "X-API-TOKEN: cv9pnurq50k1hii28630jy429g4b49viecrlj9529onpa6n0lti7yohioitvyotd0677rop5uszc0cnll03j" Check attachment exists # curl -I http://localhost:9000/attachment/cv9q94bq50k2r0s6nobg \ -H "X-API-TOKEN: cv9pnurq50k1hii28630jy429g4b49viecrlj9529onpa6n0lti7yohioitvyotd0677rop5uszc0cnll03j" Responsehttps://docs.infinilabs.com/coco-server/v0.11.0/docs/references/system/install/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/system/install/System Initialization # Initialization API # Initialize the Coco Server with an admin account and optional LLM configuration. This endpoint can only be called once — subsequent calls will return an error. Parameters # Field Type Required Description name string Yes Server name (becomes “{name}’s Coco Server”). email string Yes Admin user email address. password string Yes Admin user password.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/system/settings/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/system/settings/System Settings # System Settings API # Below is the field description for the system settings. Field Type Description server.name string The server display name. server.endpoint string The server’s public endpoint URL. server.provider.name string Provider organization name. server.provider.icon string Provider icon URL. server.provider.website string Provider website URL. server.provider.banner string Provider banner image URL.https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/file_type_detection/Mon, 01 Jan 0001 00:00:00 +0000https://docs.infinilabs.com/coco-server/v0.11.0/docs/references/pipeline_processors/document_processor/file_type_detection/File Type Detection Processor # Detects the MIME type and content category of a document based on its file extension. Documents that already have this information set are skipped. After processing, each document will carry: Field Example value Description MIME type image/png Standard MIME type Content type image The primary type portion of the MIME type Content category image, document, text A high-level category Configuration # Parameter Type Required Default Description message_field string No messages Pipeline context key for the input messages output_queue object No null Queue to push processed documents to Example # - file_type_detection: output_queue: name: "documents_typed"