Superfact

Turn any AI chat or agent run into a beautiful, shareable public page.

От сообществаРаботаетБез входаГлобальныйБесплатноТолько чтение

Что умеет

  • Superfact Publish: Publish a NEW structured artifact at a NEW URL. Pass JSON data shaped per the chosen template. ALWAYS call superfact_get_template_schema(template) first to fetch the exact field sha
  • Superfact Update Publish: Update an EXISTING Superfact at its current URL. The URL stays the same — the same link the user already shared will render the new content. Use this tool when the user says:
  • Superfact Get Publish: Fetch the current data of an existing Superfact you own. Returns the template, title, data, sources, embeds, visibility, and timestamps. Use this BEFORE superfact_update_publish

Какие данные видит

Нужен ли аккаунт

Не нужен: сервер работает без входа

Turn any AI chat or agent run into a beautiful, shareable public page.

Как подключить

Как подключить Superfact к Claude

Адрес сервера, который нужно вставить в настройках Claude:

https://superfact--superfact.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 Superfact к ChatGPT

Адрес сервера, который нужно вставить в настройках ChatGPT:

https://superfact--superfact.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 Superfact к Cursor

Адрес сервера, который нужно вставить в настройках Cursor:

https://superfact--superfact.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 (6)

Технические названия из tools/list. Нужны только разработчикам.

superfact_publishPublish a NEW structured artifact at a NEW URL. Pass JSON data shaped per the chosen template. ALWAYS call superfact_get_template_schema(template) first to fetch the exact field shape, unless you already have that template's schema in this conversation. Do not infer the shape from the template name or this description. Templates use precise field names and nested structures (e.g. evidence is a discriminated-union object, not a string; some fields are headline/body not title/summary) that are easy to guess wrong, and a wrong shape fails the call. If the user is referring to an existing Superfact URL (one you previously published in this conversation, or one they pasted), call superfact_update_publish INSTEAD. This tool always creates a new artifact at a new URL — never use it to edit an existing one. Available templates (call superfact_list_templates for the live catalog): - exec-summary: A one-page briefing with TL;DR, an optional overview strip (period, prepared by, prepared for, next checkpoint), an optional metrics row, body sections (each with an optional kind tag like 'risk' or 'progress' and optional source citations), an optional risks list, and decisions needed. Use this for weekly status updates, quarterly reviews, CEO updates, strategy summaries, or anything where the reader wants the punchline first, then the supporting detail, then the action items. Pass envelope-level sources and reference their ids from sections via source_ids to get numbered citation badges + a sources section at the bottom. - trip-plan: A trip itinerary, leisure or business, single-day or multi-day. TLDR + overview (destination, dates, travelers, style) + day-by-day timeline with items (time, title, description, optional location, optional source citations, optional tags) + optional logistics (lodging, transport, budget) + optional essentials (pack/book list). Use this for vacation plans, weekend trips, business-trip itineraries, customer-visit schedules, conference weeks, road trips, and city guides shared with a partner or team. Source citations are first-class: pass sources at the envelope level and reference their ids from items, and the renderer shows numbered badges + a sources section at the bottom. - agent-progress-log: A reverse-chronological activity feed for multi-agent workflows and team sprint coordination. Lists actors (agents and/or humans) and their entries. Each entry has a scannable one-line title plus optional markdown detail (use detail for blockers and decisions, leave it null for routine progress), a timestamp, optional blocked/done status, and an optional artifact link (PR, deploy, doc, screenshot). Use this when one or more AI agents (or people) are doing work over time and stakeholders need a single URL that stays useful as updates land. Entries are append-friendly: the renderer sorts and groups by date, and the artifact is designed to grow as agents call append_to_publish. Skips a Sources block (per-entry links carry the attribution). - research-brief: A research synthesis with TLDR, ranked findings (each with optional evidence: a verbatim quote, a stat, a link, or free prose), optional overview metadata (period, audience, method, scope), optional context, optional LLM-named blocks for industry-specific sections (methodology, assumptions, limitations, vendor landscape, etc.), optional narrative takeaway, optional recommendations, and optional open questions. Use for user research write-ups, competitive analysis, market sizing, technical evaluations, secondary research synthesis, and anything else that ranks findings with evidence. Pass envelope-level sources and reference their ids from findings via source_ids to get numbered citation badges + a sources section at the bottom. Style: do not use em-dashes (—) anywhere in `data`. Use periods, commas, or colons instead. Em-dashes are the single strongest typographic tell of LLM-generated content and undercut the artifact's hand-authored feel. Always announce your template choice in one sentence and tell the user they can republish in a different template later. Returns a sharable URL on success.
superfact_update_publishUpdate an EXISTING Superfact at its current URL. The URL stays the same — the same link the user already shared will render the new content. Use this tool when the user says: - "update the…", "change the…", "edit…", "revise…", "fix the typo in…", "add a section about…" - or otherwise refers to an existing Superfact URL they want modified NEVER use this tool to create a new artifact — call superfact_publish for that. Identify the publish by passing EITHER `url` (the full https://…/s/<uuid> URL, what the user usually has) OR `id` (the bare UUID). The publish must belong to your workspace. `data` is a FULL REPLACEMENT — pass the complete new data object, not a diff. If you don't already have the current data in your context (e.g. you're in a fresh conversation and the user just pasted a URL), call superfact_get_publish first to fetch it, then mutate, then call this. If you are unsure of the template's exact field shape, call superfact_get_template_schema(template) first rather than guessing — a wrong shape fails the call. Style: do not use em-dashes (—) anywhere in `data`. Use periods, commas, or colons instead. Em-dashes are the single strongest typographic tell of LLM-generated content and undercut the artifact's hand-authored feel. Returns the same URL plus the new updated_at timestamp.
superfact_get_publishFetch the current data of an existing Superfact you own. Returns the template, title, data, sources, embeds, visibility, and timestamps. Use this BEFORE superfact_update_publish when you don't already have the artifact's data in your conversation context — for example, a new conversation where the user pasted an existing URL. Fetch, mutate the field the user asked about, then call update_publish with the full new data. Scope: returns artifacts in your workspace only. If you want to view someone else's public Superfact, open the /s/<uuid> URL directly — this tool doesn't expose raw structured data for artifacts owned by other workspaces. Identify the publish by passing EITHER `url` (full URL, what users usually have) OR `id` (the bare UUID).
superfact_list_templatesList the available Superfact templates with their slugs, names, and descriptions. Call this first if the user did not specify a template by name.
superfact_get_template_schemaFetch the JSON Schema for a template's data. Call this BEFORE superfact_publish or superfact_update_publish to get the exact field shape, unless you already have it in this conversation. The schema is the source of truth for field names, which fields are required, and nested shapes like the evidence discriminated union — do not infer the shape from the template name.
superfact_searchSearch the workspace's published superfacts and get back the most relevant ones with the exact passages that matched. Use this when the user asks "what do you know about X", "find our superfact on Y", "what have we published about Z", "have we written anything on …", or when you want grounding from the team's existing work before drafting a new superfact. Returns ranked superfacts, each with its title, template, URL, and the specific matching passages. Read the passages and synthesize the answer yourself, citing the superfact URLs. If you need a result's full structured data, follow up with superfact_get_publish using its id or url. Scope: only superfacts in your workspace. Hybrid semantic + keyword search, so paraphrased questions and exact-term lookups both work.
Superfact: подключить к Claude, ChatGPT, Cursor · Connectors.fun