BuiltOnRails

Stack intelligence from real production Rails apps. Compare gems, scan a Gemfile.

我该使用它吗

质量与安全性

A
描述质量
100%
模式完整度
94%
命名质量
91%
投毒风险
100%
权限匹配度
100%
协议合规性
100%

基于对工具定义和协议合规性的自动分析。

上下文开销

~4,422token 数(工具定义)
~1.6 KB典型响应大小
对注意力有显著影响(占 128k 上下文窗口的 3.45%)

这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高,可用于其他任务的注意力就越少。

安装

一键安装

将以下内容添加到你的 `claude_desktop_config.json` 文件中:

{
  "mcpServers": {
    "builtonrails": {
      "url": "https://builtonrails.com/mcp/sse"
    }
  }
}

远程端点

https://builtonrails.com/mcp/ssesse

它能做什么

工具清单

工具(18)

🟢 只读🟡 写入🔴 删除⚪ 未知
🟡submit_product(name, url, description, product_type, stacks, ...)

Add a Rails application to BuiltOnRails so its stack becomes part of the data. The product is created with 'pending' status and goes through admin approval. Attach stacks by name, or paste your Gemfile / package.json / importmap.rb content and the real stack is detected and attached automatically - the recommended path.

输入模式

{
  "type": "object",
  "properties": {
    "name": {
      "type": "string",
      "minLength": 1,
      "description": "Product name (e.g. 'Shopify')"
    },
    "url": {
      "type": "string",
      "minLength": 1,
      "description": "Product website URL (e.g. 'https://shopify.com')"
    },
    "description": {
      "type": "string",
      "minLength": 1,
      "description": "Short description of the product"
    },
    "product_type": {
      "type": "string",
      "minLength": 1,
      "description": "Type: SaaS, Gem, Open Source, Marketplace, Tool, Platform, Agency, Studio"
    },
    "stacks": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Stack names to associate (e.g. ['PostgreSQL', 'Sidekiq'])"
    },
    "gemfile": {
      "type": "string",
      "minLength": 1,
      "description": "Raw Gemfile content - stacks are auto-detected and attached"
    },
    "package_json": {
      "type": "string",
      "minLength": 1,
      "description": "Raw package.json content - stacks are auto-detected and attached"
    },
    "importmap": {
      "type": "string",
      "minLength": 1,
      "description": "Raw config/importmap.rb content - stacks are auto-detected and attached"
    }
  },
  "required": [
    "name",
    "url",
    "description"
  ]
}
🟢search_stacks(query, category)

Search for technology stacks (gems, frameworks, services) used by Rails applications. Returns matching stacks with real adoption data including product counts, GitHub stars, and category. Available categories: Framework, Database, Frontend, Background, Search, Auth, Payments, DevOps, Testing, Analytics, LLM, AI/ML, Observability, Security, Email.

输入模式

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "minLength": 1,
      "description": "Search term (e.g. 'sidekiq', 'auth', 'postgres')"
    },
    "category": {
      "type": "string",
      "minLength": 1,
      "description": "Filter by category: Framework, Database, Frontend, Background, Search, Auth, Payments, DevOps, Testing, Analytics, LLM, AI/ML, Observability, Security, Email"
    }
  },
  "required": [
    "query"
  ]
}
🟢search_products(query, type, sort, limit)

Search real production Rails applications and the stacks they actually run. Returns approved products with name, URL, type, votes, and stacks. Filter by product type or sort by votes/newest.

输入模式

{
  "type": "object",
  "properties": {
    "query": {
      "type": "string",
      "minLength": 1,
      "description": "Search term to match product name (e.g. 'shopify', 'hey')"
    },
    "type": {
      "type": "string",
      "minLength": 1,
      "description": "Filter by product type (e.g. 'SaaS', 'Gem', 'Open Source', 'Marketplace')"
    },
    "sort": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "votes",
        "newest"
      ],
      "description": "Sort order: 'votes' (most upvoted) or 'newest' (default)"
    },
    "limit": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "Max results (default: 10, max: 25)"
    }
  },
  "required": []
}
🟡my_product(action, slug, description, url, product_type, ...)

OWNERS ONLY - manage your own products over MCP. Requires a write-scope Bearer token; you can only touch products you submitted or founded. Actions: 'list' shows your products; 'update' edits fields on one of your products (description, url, product_type, tags, github_repo_path); 'post_update' publishes a product update (title + update_type, optional content/url/version) shown on your product page; 'updates' lists a product's recent updates. Params: action (required); slug (required except for 'list').

输入模式

{
  "type": "object",
  "properties": {
    "action": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "list",
        "update",
        "post_update",
        "updates"
      ],
      "description": "One of: list, update, post_update, updates"
    },
    "slug": {
      "type": "string",
      "minLength": 1,
      "description": "Product slug — required for update/post_update/updates"
    },
    "description": {
      "type": "string",
      "minLength": 1,
      "description": "update: new product description"
    },
    "url": {
      "type": "string",
      "minLength": 1,
      "description": "update: new product URL / post_update: link for the update"
    },
    "product_type": {
      "type": "string",
      "minLength": 1,
      "description": "update: SaaS, Gem, Open Source, Marketplace, Tool, Platform, Agency, Studio"
    },
    "tags": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "update: replace product tags"
    },
    "github_repo_path": {
      "type": "string",
      "minLength": 1,
      "description": "update: GitHub repo path (e.g. 'rails/rails')"
    },
    "title": {
      "type": "string",
      "minLength": 1,
      "description": "post_update: update title (required for post_update)"
    },
    "update_type": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "release",
        "update",
        "feature",
        "blog",
        "video",
        "announcement",
        "milestone",
        "listing",
        "episode",
        "issue"
      ],
      "description": "post_update: release, update, feature, blog, video, announcement, milestone, listing, episode, issue"
    },
    "content": {
      "type": "string",
      "minLength": 1,
      "description": "post_update: update body (markdown, max 10k chars)"
    },
    "version": {
      "type": "string",
      "minLength": 1,
      "description": "post_update: version string for 'release' updates (e.g. '2.1.0')"
    }
  },
  "required": [
    "action"
  ]
}
🟢get_stack_trends(stack_name, days)

Get adoption trend data for a technology stack over time. Returns historical usage snapshots showing how many Rails apps use the stack, plus growth percentage. Useful for understanding if a stack is growing or declining.

输入模式

{
  "type": "object",
  "properties": {
    "stack_name": {
      "type": "string",
      "minLength": 1,
      "description": "Stack name (e.g. 'Solid Queue', 'Sidekiq')"
    },
    "days": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "Number of days of trend data (default: 90, max: 365)"
    }
  },
  "required": [
    "stack_name"
  ]
}
🟢get_stack_recommendations(current_stacks, category)

Get stack recommendations based on what real Rails apps use together. Given your current stacks, recommends the most popular stacks in a target category based on co-occurrence in production Rails applications. Available categories: Framework, Database, Frontend, Background, Search, Auth, Payments, DevOps, Testing, Analytics, LLM, AI/ML, Observability, Security, Email.

输入模式

{
  "type": "object",
  "properties": {
    "current_stacks": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Your current stacks (e.g. ['PostgreSQL', 'Hotwire', 'Tailwind CSS'])"
    },
    "category": {
      "type": "string",
      "minLength": 1,
      "description": "Category to get recommendations for (e.g. 'Background', 'Auth', 'Testing')"
    }
  },
  "required": [
    "current_stacks",
    "category"
  ]
}
🟢get_products_using_stacks(stacks, limit)

Find real Rails applications that use specific technology stacks. Returns approved products using ALL specified stacks, with their full stack lists and vote counts. Great for finding real-world examples of specific tech stack combinations.

输入模式

{
  "type": "object",
  "properties": {
    "stacks": {
      "type": "array",
      "items": {
        "type": "string"
      },
      "description": "Stack names to search for (e.g. ['PostgreSQL', 'Sidekiq']). Returns products using ALL specified stacks."
    },
    "limit": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "Maximum number of products to return (default: 10, max: 25)"
    }
  },
  "required": [
    "stacks"
  ]
}
🟢get_product(slug)

Get detailed information about a specific production Rails application. Returns full product details including description, tags, member, stacks, and vote count.

输入模式

{
  "type": "object",
  "properties": {
    "slug": {
      "type": "string",
      "minLength": 1,
      "description": "Product slug (e.g. 'shopify', 'hey', 'github')"
    }
  },
  "required": [
    "slug"
  ]
}
🟢get_news(category, date, query, limit)

Read curated Rails ecosystem news and articles - the same pipeline behind builtonrails.com/daily. Returns recent items (title, url, summary, category, source). Filter by category (article, release, podcast, video, news), a specific date (YYYY-MM-DD), or a search term. Use this for 'what's new in Rails', release announcements, and article discovery.

输入模式

{
  "type": "object",
  "properties": {
    "category": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "article",
        "release",
        "podcast",
        "video",
        "news"
      ],
      "description": "Filter: article, release, podcast, video, news"
    },
    "date": {
      "type": "string",
      "minLength": 1,
      "description": "Specific day, YYYY-MM-DD"
    },
    "query": {
      "type": "string",
      "minLength": 1,
      "description": "Search term matched against title and summary"
    },
    "limit": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "Max results (default 15, max 50)"
    }
  },
  "required": []
}
🟢get_collection(slug)

Get a curated stack collection by slug. Collections are hand-picked sets of technology stacks with notes and optional LLM prompts for AI context. Returns the collection with all stacks in order, notes, and the LLM prompt if available.

输入模式

{
  "type": "object",
  "properties": {
    "slug": {
      "type": "string",
      "minLength": 1,
      "description": "Collection slug (e.g. 'modern-rails-2024', 'essential-saas-stack')"
    }
  },
  "required": [
    "slug"
  ]
}
🟢get_categories

Get all available stack categories with descriptions and stack counts. Useful for understanding what types of technologies are tracked and filtering searches.

输入模式

{
  "type": "object",
  "properties": {},
  "required": []
}
🟢detect_stacks(gemfile, gemspec, package_json, importmap)

Detect technology stacks from a Gemfile, gemspec, package.json, or importmap.rb content. Paste your dependency file content and get back matched stacks with real adoption data. The killer feature for coding agents — understand your project's stack landscape instantly.

输入模式

{
  "type": "object",
  "properties": {
    "gemfile": {
      "type": "string",
      "minLength": 1,
      "description": "Raw Gemfile content"
    },
    "gemspec": {
      "type": "string",
      "minLength": 1,
      "description": "Raw .gemspec content — gems declare deps here, not in the Gemfile"
    },
    "package_json": {
      "type": "string",
      "minLength": 1,
      "description": "Raw package.json content"
    },
    "importmap": {
      "type": "string",
      "minLength": 1,
      "description": "Raw config/importmap.rb content"
    }
  },
  "required": []
}
🟢compare_stacks(stack1, stack2)

Compare two technology stacks head-to-head with real adoption data. Returns products count, GitHub stars, collections, growth trends, and a winner. Uses the same scoring algorithm as the BuiltOnRails Stack Showdown feature.

输入模式

{
  "type": "object",
  "properties": {
    "stack1": {
      "type": "string",
      "minLength": 1,
      "description": "First stack name (e.g. 'Sidekiq')"
    },
    "stack2": {
      "type": "string",
      "minLength": 1,
      "description": "Second stack name (e.g. 'Solid Queue')"
    }
  },
  "required": [
    "stack1",
    "stack2"
  ]
}
⚪admin_sponsors(action, id, name, url, position, ...)

ADMIN ONLY - manage site sponsors over MCP. Requires an admin Bearer token. Actions: 'list' shows sponsors with performance (impressions, clicks, CTR, running state); 'create' adds a manual sponsor (name, url, position, optional tagline/dates); 'update' changes fields on a sponsor by id; 'activate' / 'deactivate' toggle a sponsor; 'inquiries' lists sponsorship inquiries; 'inquiry_status' moves an inquiry through pending -> contacted -> converted/declined; 'orders' lists self-serve sponsor orders.

输入模式

{
  "type": "object",
  "properties": {
    "action": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "list",
        "create",
        "update",
        "activate",
        "deactivate",
        "inquiries",
        "inquiry_status",
        "orders"
      ],
      "description": "list | create | update | activate | deactivate | inquiries | inquiry_status | orders"
    },
    "id": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "Sponsor id (update/activate/deactivate) or inquiry id (inquiry_status)"
    },
    "name": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: sponsor name"
    },
    "url": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: sponsor link URL"
    },
    "position": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "sidebar",
        "banner",
        "featured"
      ],
      "description": "create/update: sidebar | banner | featured"
    },
    "tagline": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: short tagline shown on the ad"
    },
    "cta_text": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: call-to-action label (default 'Learn More')"
    },
    "starts_at": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: start date, YYYY-MM-DD (omit = immediately)"
    },
    "ends_at": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: end date, YYYY-MM-DD (omit = open-ended)"
    },
    "status": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "pending",
        "contacted",
        "converted",
        "declined"
      ],
      "description": "inquiry_status: pending | contacted | converted | declined"
    }
  },
  "required": [
    "action"
  ]
}
🟡admin_newsletters(action, id, subject, content, segment)

ADMIN ONLY - queue and send promotional/update newsletters over MCP. Requires an admin Bearer token. Each newsletter is only ever delivered once per subscriber, even across retries. The 'all_subscribed' segment is additionally throttled: it skips anyone who got a newsletter in the last 7 days or any email at all in the last 2 days, so promos never stack on top of other mail. Leaving an 'all_subscribed' newsletter as a draft (don't call send_now) queues it for the automatic weekly digest every Monday 10am, which folds all pending drafts into one combined email instead of several separate sends. Actions: 'list' shows recent newsletters with status and segment; 'segments' returns live, throttle-aware recipient counts for each segment so you can preview reach before sending; 'pending_digest' lists drafts currently queued for next Monday's digest; 'create' drafts a newsletter (subject, content, segment); 'update' edits a draft/scheduled newsletter's fields; 'send_now' sends immediately in the background instead of waiting for the digest; 'destroy' deletes a newsletter.

输入模式

{
  "type": "object",
  "properties": {
    "action": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "list",
        "segments",
        "pending_digest",
        "create",
        "update",
        "send_now",
        "destroy"
      ],
      "description": "list | segments | pending_digest | create | update | send_now | destroy"
    },
    "id": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "Newsletter id - required for update/send_now/destroy"
    },
    "subject": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: email subject line"
    },
    "content": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: body, plain text or markdown"
    },
    "segment": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "inactive",
        "all_subscribed"
      ],
      "description": "create/update: 'inactive' (re-engage users quiet 14+ days) or 'all_subscribed' (promotional blast to everyone opted into the newsletter). Defaults to 'inactive'."
    }
  },
  "required": [
    "action"
  ]
}
🟢admin_news(action, ids, feed, query, limit)

ADMIN ONLY - batch-review the curated news queue (ContentItems behind /daily and the newsletter) over MCP. Requires an admin Bearer token. Actions: 'queue' lists items awaiting review (filter by feed name or search term); 'select' / 'reject' batch-review by ids (array); 'stats' returns queue counts by status and top feeds. Typical flow: queue -> read titles -> select the good ids, reject the noise, in batches.

输入模式

{
  "type": "object",
  "properties": {
    "action": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "queue",
        "select",
        "reject",
        "stats"
      ],
      "description": "queue | select | reject | stats"
    },
    "ids": {
      "type": "array",
      "items": {
        "type": "integer"
      },
      "description": "ContentItem ids - required for select/reject (max 100 per call)"
    },
    "feed": {
      "type": "string",
      "minLength": 1,
      "description": "queue: filter by content feed name (partial match)"
    },
    "query": {
      "type": "string",
      "minLength": 1,
      "description": "queue: search term matched against title"
    },
    "limit": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "queue: max results (default 25, max 50)"
    }
  },
  "required": [
    "action"
  ]
}
⚪admin_moderation(action, slug, limit)

ADMIN ONLY - moderate BuiltOnRails listings over MCP. Requires an admin Bearer token (Authorization header). Actions: 'pending' lists products awaiting review; 'approve' / 'reject' run the review lifecycle on a product slug; 'feature' / 'unfeature' toggle homepage featuring; 'stats' returns moderation counts. Params: action (required); slug (required for approve/reject/feature/unfeature); limit (pending only, default 10, max 50).

输入模式

{
  "type": "object",
  "properties": {
    "action": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "pending",
        "approve",
        "reject",
        "feature",
        "unfeature",
        "stats"
      ],
      "description": "One of: pending, approve, reject, feature, unfeature, stats"
    },
    "slug": {
      "type": "string",
      "minLength": 1,
      "description": "Product slug — required for approve/reject/feature/unfeature"
    },
    "limit": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "Max results for 'pending' (default 10, max 50)"
    }
  },
  "required": [
    "action"
  ]
}
🟢admin_changelog(action, id, title, description, category, ...)

ADMIN ONLY - the feature/changelog ledger over MCP. Requires an admin Bearer token. This is the source of truth for what shipped - log an entry here right after shipping something instead of hand-editing docs/changelog.md, so the record can't drift from reality. Entries left un-digested automatically get folded into the next Monday promotional digest (see admin_newsletters). Actions: 'list' shows recent entries (filter by category or digest status); 'create' logs a new entry (title, description, category, optional version); 'update' edits an entry by id; 'destroy' removes one.

输入模式

{
  "type": "object",
  "properties": {
    "action": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "list",
        "create",
        "update",
        "destroy"
      ],
      "description": "list | create | update | destroy"
    },
    "id": {
      "type": "integer",
      "not": {
        "type": "null"
      },
      "description": "Entry id - required for update/destroy"
    },
    "title": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: short headline"
    },
    "description": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: the technical detail"
    },
    "category": {
      "type": "string",
      "minLength": 1,
      "enum": [
        "added",
        "changed",
        "fixed",
        "improved"
      ],
      "description": "create/update: added | changed | fixed | improved"
    },
    "version": {
      "type": "string",
      "minLength": 1,
      "description": "create/update: e.g. '6.11.0' (optional)"
    },
    "pending_only": {
      "type": "boolean",
      "not": {
        "type": "null"
      },
      "description": "list: only entries not yet included in a digest"
    }
  },
  "required": [
    "action"
  ]
}

社区

评价此服务器

证据

最近观测

已验证未记录版本18 个工具
已验证未记录版本18 个工具