AlphaCreek SEC Filings MCP

ai.alphacreek由 alphacreek.ai 發布官方發布者v1.0.1更新於 Oct 24, 2026

Access SEC filings efficiently (10-K, 10-Q, etc), save time and tokens, and get cited answers.

已驗證Streamable HTTP可網頁執行金融知識與記憶

概覽

AI 產生的概覽

遠端 MCP 伺服器,用來讀取 SEC 與英國 FCA 申報文件:最新申報、文件清單、目錄、節點內容與揭露檢索。

可以做什麼
  • 依公司或股票代號取得最新申報,或依最新入庫順序列出 SEC 與 FCA 的申報
  • 依時間倒序列出某股票代號可用的申報,包含英國 FCA NSM 報告與 DEF 14A 委託書
  • 取得某份申報的完整目錄,並讀取一個或多個導覽節點的內容
  • 用概念組在美國 SEC 申報節點(8-K、6-K、10-K、10-Q、20-F)中檢索某項揭露陳述
適用情境
適合需要回答公司揭露、年報季報、委託書或英國監管公告相關問題的情境,內容直接取自申報原文,而非一般網頁搜尋。
執行需求
遠端 streamable HTTP 端點,不需本機安裝套件。需在 Authorization 標頭提供 bearer 權杖:Claude 或 ChatGPT 的 OAuth 存取權杖,或來自 AlphaCreek 主控台的 MCP API 金鑰。
安裝前請注意
Authorization 標頭攜帶金鑰或 OAuth 權杖,應避免寫入日誌或共用設定。使用需綁定 AlphaCreek 帳號,檢索請求會傳送給第三方服務商。

安裝

在 SourceWeft 中

  1. 開啟 儀表板中的 AlphaCreek SEC Filings MCP,將其新增到工作區。
  2. 連接 Authorization token 憑證。它們會被私密儲存,絕不會在公開頁面上展示。
  3. 為需要使用其工具的對話啟用該服務。

Web executable,透過 Streamable HTTP。 遠端服務在工作區中設定後即可從網頁執行環境執行。

交給代理程式安裝

把這段提示詞貼上到 Claude Code、Codex、Cursor 等代理程式,或 SourceWeft 對話中。代理程式會讀取該伺服器的安裝指南,先向你展示發布者、端點和登入方式,經你同意後再把它新增到你的客戶端。金鑰不會寫進任何命令。

閱讀 https://sourceweft.com/mcp/ai.alphacreek%2Falphacreek-mcp/install.md,並按其中說明把這個 MCP 伺服器接入我的客戶端。修改任何設定之前,先告訴我它的發布者、端點和登入方式,等我確認。不要把 API 金鑰寫進命令或我會提交的檔案;問我金鑰應從哪裡取得。

檢視代理程式所依據的安裝指南

其他 MCP 客戶端

把它新增到你客戶端的 mcpServers 設定中。

{
  "mcpServers": {
    "alphacreek-sec-filings": {
      "type": "http",
      "url": "https://mcp.alphacreek.ai/mcp",
      "headers": {
        "Authorization": "Bearer <YOUR_TOKEN>"
      }
    }
  }
}

工具

5

get_latest_filing

Get latest filing

Return latest filing metadata for a company or ticker. If ticker and company are omitted, return newly ingested filings newest-first. Omit filing_registry in that mode to include both SEC and FCA. Company results include one metadata object; newly ingested results include a filings list. Each row includes Filing metadata object. Includes `document_type` (Canonical document type. SEC: 10-K, 10-Q, 20-F, 6-K, 8-K, DEF 14A. FCA: annual, half-year, press.) and, for FCA press rows, `announcement_type` (FCA press subtype when document_type is press; null for annual and half-year. Values: 1st Quarter Results, 2nd Quarter Results, 3rd Quarter Results, 4th Quarter Results, Final Results, Preliminary Results, Share Buyback, Trading Update.).

輸入參數limittickercompanydocument_typefiling_registry

list_filings

List available filings

List available filings for a ticker (newest first) with artifact_document_id and dates. Use filing_registry=fca_nsm for UK FCA NSM reports. Use this first when you need a specific reporting period, including DEF 14A proxy statements (pass document_type=DEF 14A). Each row includes Filing metadata object. Includes `document_type` (Canonical document type. SEC: 10-K, 10-Q, 20-F, 6-K, 8-K, DEF 14A. FCA: annual, half-year, press.) and, for FCA press rows, `announcement_type` (FCA press subtype when document_type is press; null for annual and half-year. Values: 1st Quarter Results, 2nd Quarter Results, 3rd Quarter Results, 4th Quarter Results, Final Results, Preliminary Results, Share Buyback, Trading Update.).

輸入參數limittickercompanydocument_typefiling_registry

get_filing_toc

Get filing table of contents

Return the full filing table of contents for one filing. If artifact_document_id is omitted, resolves the single latest filing for ticker/company (and optional document_type) only. Use list_filings first to discover a specific reporting period. Pass document_type=DEF 14A for US proxy statements. For DEF 14A, TOC rows are native proxy section labels when the filing has a linked HTML TOC; otherwise they use sec-parser headings. For FCA filings, pass filing_registry=fca_nsm; TOC node ids are HTML page element ids (e.g. pf2a, page12, pageView1) for annual ESEF reports or sec-parser ids (e.g. sp-0000001) for half-year RNS HTML reports.

輸入參數tickercompanydocument_typefiling_registryartifact_document_id

read_node_content

Read node content

Return content for one or more navigation nodes in a filing. Pass one node via node_id or several via node_ids, using the node ids from TOC lines (the same values shown as NODE_ID lines in this tool's output). Returns plain MCP text: an ARTIFACT_DOCUMENT_ID header, then per-node blocks with NODE_ID, TITLE, CITATION_URL, CITATION_MARKDOWN, CONTENT_START … CONTENT_END. Each block's CITATION_URL links to the reader location for that node; CITATION_MARKDOWN is the same link as ready-to-paste markdown [TITLE](CITATION_URL). A requested TOC node may expand into more granular child NODE_ID blocks, each with its own TITLE, CITATION_URL, and CITATION_MARKDOWN. Responses that include citations end with a CITATIONS_IN_THIS_RESPONSE list.

輸入參數node_idnode_idsartifact_document_id必填

search_sec_event_filings

Search SEC filing nodes

Search US SEC filing nodes for a disclosure claim. The search covers 8-K, 6-K, 10-K, 10-Q, and 20-F. Omit document_type to search all of those forms. Pass document_type to search only the forms you name. Do not use this for DEF 14A. How to search: 1. Always pass concept_groups. Every group is required (AND). Within each group's any_of list, one alternative must match (OR). Use 2-5 groups for separate parts of the claim: topic, action, metric, and a causal link when that link is essential. Example: geography + weaker demand + decline, or AI + productivity + margin. Add synonyms. English stemming covers inflections (decline/declined), not synonyms (sales is not revenue). Do not add a group for a detail the filing may leave implicit. 2. query is optional. Each item is an exact adjacent-token phrase. Put alternate full phrasings in the same list. Query plus concept_groups is hybrid search: an exact phrase ranks first, so a generic phrase such as "margin expansion" can outrank the causal sentence you want. Prefer phrases that state the relationship. Do not pass a broad topic word alone. Do not search with query only. Omit query for concept-only search. 3. start_date and end_date are required. tickers limits issuers. Optional document_type limits the search to one or more of 8-K, 6-K, 10-K, 10-Q, and 20-F. Omit it to search all of those forms. How to continue: 4. Each call returns one page. There is no total count. If HAS_MORE is true, call again with cursor set to NEXT_CURSOR and the same concept_groups, dates, query, tickers, ne_tickers, document_type, and match_scope. A cursor from different inputs is rejected. An empty page with HAS_MORE true is not the end: call again with that cursor. Stop when HAS_MORE is false or when later pages are not relevant. 5. If a page is mostly co-occurrence (the terms appear, but not the claim), start a new search with no cursor. Add a concept group or exact causal phrases. If a page is too sparse, start a new search with no cursor and d

輸入參數limitquerycursortickersend_date必填ne_tickersstart_date必填match_scopedocument_typeconcept_groups必填

版本歷史

1
  1. v1.0.1最新Sep 16, 2026

評分

已滿足適用檢查項的 83%

該評分衡量此收錄條目的完整程度,依據註冊中心條目、端點與公開來源計算得出。它不是對伺服器本身的評價,也不表示伺服器安全或目前可用:可連通性是帶日期的檢查結果,評分也不包含安裝量或使用者評價。

  • 近 30 天內可連通必需25 分
  • 已宣告認證方式必需10 分
  • 已列出工具必需15 分
  • 描述必需15 分
  • 圖示10 分
  • 文件15 分

    請宣告包含 README 的 GitHub 儲存庫,或在 OAuth 受保護資源後設資料中提供 resource_documentation。

  • 授權條款不適用

    託管服務按其提供方的條款使用,因此這裡不要求授權條款。此檢查項不計入百分比。