CNAPS.ai MCP Server

CNAPS.ai MCP Server: собирает AI-пайплайны из 70+ моделей по описанию задачи на языке

От сообществаРаботаетБез входаГлобальныйБесплатноТолько чтение

Что умеет

  • Проверять совместимость моделей перед сборкой пайплайна
  • Создавать и запускать многошаговый пайплайн за один вызов
  • Отменять или дублировать выполняющийся флоу
  • Диагностировать ошибки выполнения и запускать тестовый прогон
  • Смотреть статус задачи, использование квоты и уведомления

Какие данные видит

Ассистент работает от имени аккаунта пользователя CNAPS Studio через OAuth 2.1, видит его пайплайны, модели и статистику использования.

Нужен ли аккаунт

Не нужен: сервер работает без входа

CNAPS.ai MCP Server подключает ассистента к CNAPS Studio — no-code платформе для AI-пайплайнов. По описанию задачи вроде «увеличь фото в 4 раза» ассистент выбирает из более 70 моделей и 50 инструментов обработки, собирает цепочку и запускает её. Работает через OAuth 2.1, совместим с Claude, n8n и Zapier через их MCP-клиенты, тариф фиксированный без оплаты за каждый вызов модели.

Как подключить

Как подключить CNAPS.ai MCP Server к Claude

Адрес сервера, который нужно вставить в настройках Claude:

https://cnaps-studio.run.tools

  1. Откройте Claude (сайт claude.ai или приложение для компьютера).
  2. Перейдите в Настройки → Коннекторы (в английском интерфейсе: Settings → Connectors).
  3. Нажмите Добавить пользовательский коннектор (Add custom connector).
  4. Вставьте адрес сервера, скопированный ниже, в поле URL удалённого MCP-сервера и нажмите Добавить.
  5. В чате нажмите + → Коннекторы и включите новый коннектор.

Пользовательские коннекторы доступны на тарифах Free, Pro, Max, Team и Enterprise. На Free можно добавить только один. В Team и Enterprise коннектор сначала добавляет владелец организации в Настройки организации → Коннекторы.

Авторизация

После нажатия Добавить откроется окно входа в сервис. Войдите под своим аккаунтом и подтвердите доступ. Пароль Claude не видит.

Как подключить CNAPS.ai MCP Server к ChatGPT

Адрес сервера, который нужно вставить в настройках ChatGPT:

https://cnaps-studio.run.tools

  1. Откройте ChatGPT в браузере (chatgpt.com). Нужен тариф Plus, Pro, Business, Enterprise или Edu.
  2. Один раз включите режим разработчика: Настройки → Приложения → Дополнительные настройки → Режим разработчика (Settings → Apps → Advanced settings → Developer mode).
  3. Откройте Настройки → Коннекторы и нажмите Создать (Create).
  4. Заполните форму: Название (любое), Описание (одна фраза, что делает сервис), URL MCP-сервера (скопируйте ниже).
  5. В поле Аутентификация выберите OAuth, если сервис требует вход, иначе Нет. Нажмите Создать.
  6. В новом чате откройте + → Приложения/Коннекторы и включите коннектор.

Название раздела у OpenAI менялось (Connectors → Apps/Plugins). Если пункт называется иначе, ищите «режим разработчика» в настройках. В рабочих пространствах Business/Enterprise администратор должен разрешить пользовательские коннекторы.

Авторизация

При первом использовании ChatGPT откроет окно входа в сервис. Войдите и подтвердите доступ.

Как подключить CNAPS.ai MCP Server к Cursor

Адрес сервера, который нужно вставить в настройках Cursor:

https://cnaps-studio.run.tools

Быстрый способ: нажмите кнопку Открыть в Cursor ниже, Cursor сам предложит добавить сервер.

Вручную:

  1. В Cursor откройте Settings → Cursor Settings → MCP и нажмите Add new global MCP server.
  2. Вставьте JSON, скопированный ниже, в открывшийся файл ~/.cursor/mcp.json (для одного проекта: .cursor/mcp.json в корне проекта).
  3. Сохраните файл. Сервер появится в списке MCP; при необходимости пройдите авторизацию по кнопке рядом с ним.

Авторизация

Если сервис требует вход, рядом с сервером в списке MCP появится кнопка авторизации.

Список инструментов сервера (45)

Технические названия из tools/list. Нужны только разработчикам.

cancel_flowCancel a currently running workflow. Sends a stop signal — the flow will stop after completing the current processing step. Use get_flow_status to confirm it has stopped.
check_pipeline_compatibilityCheck whether a sequence of AI models can be wired into a valid pipeline BEFORE creating a flow. Returns compatibility status, wiring details, required inputs, and final outputs. Use this to validate multi-model pipelines and avoid create_flow errors.
create_and_run_pipelineCreate AND immediately run a multi-model pipeline in one step. Combines create_pipeline + run_flow — no need to parse run_hint or call run_flow separately. After calling this, just poll get_flow_status for completion. Use check_pipeline_compatibility first to verify models can be wired together.
create_community_postPublish one of your flows to the cnaps.ai community. Creates a community post with a snapshot of the flow. Others can then discover and fork your workflow. Requires a source_flow_id (from list_flows) and a title.
create_flowCreate a new AI workflow. Just provide ai_model nodes in pipeline order — input/output nodes and connections are auto-generated. For multi-model pipelines, models are wired by matching output→input port types (e.g., model A's image output connects to model B's image input). The original input is shared (fan-out) when multiple models need it. Optionally provide input_files to pre-populate input nodes.
create_pipelineCreate a multi-model pipeline flow from a simple list of model IDs. This is the EASIEST way to create flows — just provide model IDs in pipeline order. Input/output nodes and connections are auto-generated by type matching. Use check_pipeline_compatibility first to verify models can be wired together. For advanced cases (custom connections, mixed node types), use create_flow instead.
delete_flowMove a workflow to trash. The flow can be restored later from the trash.
describe_flowGet a human-readable description of what a workflow does. Returns the processing pipeline in natural language: input types, model chain with descriptions, and output types. Use this when the user asks 'what does this flow do?' instead of returning raw JSON from get_flow.
diagnose_errorAnalyze a failed pipeline execution and suggest fixes. Provide the flow_id of a failed flow. Returns root cause analysis, the specific failing node, a concrete fix suggestion (e.g. insert a Resize node, change a parameter), and confidence level. Use this whenever get_flow_status shows a failed node.
dry_run_flowPreview what a flow would look like WITHOUT actually creating it. Runs the auto-scaffold logic (input/output node generation, port-based wiring) and returns the full node structure, connections, and run_hint. Use this to verify the pipeline structure before committing to create_flow or create_pipeline. Only accepts ai_model nodes — input/output nodes are auto-generated.
duplicate_flowDuplicate an existing workflow. Creates a deep copy with fresh state (no results). Use this to copy a flow and then modify it with update_flow, instead of recreating from scratch.
find_compatible_modelsFind models that can connect to a given model in a pipeline. Use 'downstream' to find models whose inputs match this model's outputs (what can come AFTER). Use 'upstream' to find models whose outputs match this model's inputs (what can come BEFORE). Returns compatible models sorted by number of matching ports. Use this BEFORE creating a pipeline to ensure models can be wired together.
fork_community_flowFork (copy) a community workflow into your workspace. After forking, the flow is yours — you can run it with run_flow, modify it, or use it as a starting point. Get post_id from search_community results.
fork_templateFork (copy) a workflow template into your workspace. After forking, you can run it with run_flow.
get_batch_statusCheck the status of a batch run. Returns progress and output file URLs when completed.
get_community_postGet detailed information about a community post, including title, content, tags, input/output types, and engagement stats. Use this to preview a post before forking it.
get_flowGet detailed information about a specific workflow, including its nodes, connections, and parameters.
get_flow_statusCheck the execution status of a workflow. Returns per-node status and output file URLs when completed.
get_meGet your user profile information (ID, email, name).
get_modelGet detailed information about a specific AI model, including its inputs, outputs, parameters, and plan requirements.
get_model_parametersGet the configurable parameters schema for a specific AI model. Use this to understand what settings can be adjusted when creating a flow with this model.
get_notificationsGet your notifications (likes, comments, forks, follows, mentions). Returns cursor-paginated results. Use next_cursor from response for subsequent pages.
get_unread_notification_countGet the number of unread notifications.
get_usageGet GPU usage for your workspace. Shows usage percentage, plan limits, and reset schedule.
get_workspace_infoGet information about your current workspace, including subscription plan, member count, and seats.
list_flowsList AI workflows in your workspace. Returns flow names, statuses, and metadata with pagination.
list_llm_providersList available LLM providers and whether your workspace has configured API keys for them. Configured providers enable external model nodes in workflows.
list_modelsList available AI models on cnaps.ai with optional filters. Returns model names, types, input/output port types, and accessibility. Use filters to narrow results instead of browsing all models.
list_templatesList curated workflow templates on cnaps.ai. Templates are pre-built, tested workflows for common tasks. Fork a template with fork_template to use it.
mark_all_notifications_readMark all notifications as read.
mark_notification_readMark a specific notification as read.
optimize_flowAnalyze a flow for performance and cost optimization opportunities. Returns rule-based suggestions such as moving upscale nodes to the end of the pipeline, avoiding resolution overflow, removing redundant processing, and choosing better-performing models. Each suggestion includes a mechanical patch so the caller can apply it automatically. Use this before running a flow or while iterating on its design. Set include_llm_analysis=true to also ask Haiku for complex-pattern refinements.
rename_flowRename a workflow.
report_issueReport a tool failure or confusion to the cnaps.ai team. Use this when you encounter persistent errors, unclear documentation, or missing features. Reports are stored and analyzed to improve the platform.
restore_flowRestore a workflow from trash back to active state. Use list_flows with state='trashed' to find trashed flows.
run_batchRun a workflow in batch mode with multiple input file sets. Each input node receives an array of file URLs (uploaded via upload_file). Returns a batch_run_id to track progress.
run_flowExecute an AI workflow. The flow runs asynchronously — use get_flow_status to poll for completion. For flows with input nodes, provide file URLs (from upload_file) via the inputs parameter.
search_communitySearch community-shared AI workflows on cnaps.ai. Find workflows by keyword, input/output modality, or tags. Use this to discover existing workflows before creating new ones. Returns post IDs that can be forked with fork_community_flow.
suggest_modelsSuggest individual AI models for a specific task. Use suggest_pipeline instead when the user describes a multi-step goal. Use this only when the user specifically asks about model options for a single processing step. Returns up to 5 model recommendations with match reasons.
suggest_pipelineRECOMMENDED FIRST STEP for any image/video/text processing request. Understands natural language (English, Korean, and more) and returns the optimal model pipeline with parameter recommendations and wiring. Use this BEFORE list_models or suggest_models — it handles multi-step tasks, parameter inference (e.g. '4K' → scale_factor=4), and model selection automatically. After getting a suggestion, use create_pipeline or create_and_run_pipeline to build it.
update_flowUpdate node parameters in an existing workflow. Use this to modify model settings (e.g., scale_factor, steps, seed) without recreating the flow. Parameters are merged with existing values — only specify the parameters you want to change. Cannot update a flow that is currently running.
update_flow_tagsUpdate the tags on a workflow. Replaces all existing manual tags.
upload_fileUpload a file from a public URL for use in AI workflows. Supported types: images (png, jpg, webp, RAW), video (mp4, mov, webm, mpeg, avi, flv, wmv, 3gp), audio (wav, mp3, m4a, aac, ogg, flac), text (txt). Use this to register a video for the Video Analysis (llm-gemini-video) node from a public URL. IMPORTANT: If the user wants to upload an image/video from the chat, use upload_image instead. This tool is ONLY for when the user provides a public HTTP/HTTPS URL.
upload_imageDEFAULT method to get files from users (images, video, or audio). Use this whenever the user wants to process/upload a file but has NOT provided a direct URL or local file path. You CANNOT transfer files from the chat to cnaps.ai — even if you see an image or video in the conversation, you cannot send its bytes through MCP. This tool creates a browser upload link where the user can drag-and-drop their file (e.g. an mp4 clip for the Video Analysis node). Step 1: Call WITHOUT token → get upload link → show link to user. Step 2: IMMEDIATELY call WITH the returned token → tool waits until the user finishes uploading (up to 15 minutes) and returns file URLs. Returns URLs usable in create_flow input_files or run_flow inputs.
validate_flowValidate a workflow before running it. Checks for common issues: trashed state, missing input files, disconnected nodes, currently running. Returns {valid: true/false, issues: [...]}. Use this before run_flow to catch problems early.