Finance Toolkit
500+ transparent financial methods calculated from raw statements, not third-party endpoints.
作者提供的说明(英文原文)
SEC filings, dilution, insider & institutional ownership, and political-trade data for AI agents.
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
SIGNAL8_API_KEY, Authorization标为机密的值是你账户的钥匙。千万不要分享给别人,也不要粘贴到聊天里。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
这个地址需要在请求的 header 中放入密钥,而添加连接器的菜单无法设置。请改用 Claude Code、Cursor 或 VS Code。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"signal8": {
"command": "npx",
"args": [
"-y",
"@signal8ai/mcp"
],
"env": {
"SIGNAL8_API_KEY": "your-value"
}
}
}
}在终端中运行这条命令
claude mcp add --transport http signal8 https://mcp.signal8.ai/mcp --header 'Authorization: your-value'claude mcp add signal8 --env SIGNAL8_API_KEY=your-value -- npx -y @signal8ai/mcp运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"signal8": {
"url": "https://mcp.signal8.ai/mcp",
"headers": {
"Authorization": "your-value"
}
}
}
}粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"signal8": {
"command": "npx",
"args": [
"-y",
"@signal8ai/mcp"
],
"env": {
"SIGNAL8_API_KEY": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"signal8": {
"type": "http",
"url": "https://mcp.signal8.ai/mcp",
"headers": {
"Authorization": "your-value"
}
}
}
}粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"signal8": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@signal8ai/mcp"
],
"env": {
"SIGNAL8_API_KEY": "your-value"
}
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
这个地址需要在请求的 header 中放入密钥,而添加连接器的菜单无法设置。请改用 Claude Code、Cursor 或 VS Code。
你的选择通过 Cookie 保存在这台设备上,无需账户。
Authorization 机密 HTTP header这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
我们还没有读取这个服务器的工具列表;我们会轮流检查各个端点。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日
注册名称已通过域名 signal8.ai 验证(注册中心会核实域名所有权)
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
尚未检查;我们会轮流检查各个端点
尚未检查;我们会轮流检查各个端点
尚未检查;我们会在检查端点时读取工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
你需要填写:SIGNAL8_API_KEY, Authorization
两种方式都可以:安装在你的电脑上,或者连接云端版本。
GitHub 仓库最后更新于–,最新版本(0.19.0)于19天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。
同一分类中有:Finance Toolkit, SolidInvoice, Korea Tax Calculator
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
500+ transparent financial methods calculated from raw statements, not third-party endpoints.
Create and send invoices and quotes, manage clients, and track payments in SolidInvoice.
Korean tax calculation and tax-audit judgment MCP: 96 tools with statutory citations.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
MCP (Model Context Protocol) server for Signal8 -- AI-extracted SEC filing intelligence data.
Gives AI agents like Claude, Cursor, and other MCP-compatible tools direct access to:
npm install -g @signal8ai/mcp
npx @signal8ai/mcp
API and MCP access are included with Signal8 Pro ($59/mo, or $588/yr). A free Signal8 account cannot create API keys. Upgrade at signal8.ai/pro, then create a key in the developer portal at signal8.ai/developer/api-keys.
Using Claude.ai or Claude Desktop? You can skip the key entirely: add the hosted server as a custom connector and sign in with your Signal8 account (see Hosted Endpoint).
Add to your Claude Desktop config file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": { "signal8": { "command": "npx", "args": ["-y", "@signal8ai/mcp"], "env": { "SIGNAL8_API_KEY": "sk_live_your_key_here" } } } }
Add to .cursor/mcp.json in your project root:
{
"mcpServers": { "signal8": { "command": "npx", "args": ["-y", "@signal8ai/mcp"], "env": { "SIGNAL8_API_KEY": "sk_live_your_key_here" } } } }
Set the SIGNAL8_API_KEY environment variable and run:
SIGNAL8_API_KEY=sk_live_your_key_here npx @signal8ai/mcp
The server exposes 104 live tools at runtime. The reference tables below are a representative catalog grouped by domain, not an exhaustive list; the Extractions section documents tools that are currently disabled and not registered live. The authoritative list is what tools/list returns.
| Tool | Description |
|---|---|
search_companies | Search companies by name or ticker |
get_company_profile | Full company profile (description, market cap, sector) |
get_company_bundle | Consolidated data bundle in one call |
| Tool | Description |
|---|---|
get_quote | Real-time stock quote (price, change, volume) |
get_market_metrics | Volume averages, volatility, SMAs, trend direction |
get_short_interest | Short interest data and days-to-cover |
get_float | Float composition, shares outstanding, ownership % |
get_financials | Income statement, balance sheet, cash flow |
get_earnings | Earnings history with EPS surprises |
get_executives | Executive team roster and compensation |
get_peers | Peer/comparable companies |
get_transcripts | Earnings call transcripts |
get_news | Latest company news headlines |
get_analyst_consensus | Analyst ratings and price targets |
get_material_events | Material corporate events (8-K filings) |
get_clinical_trials | Clinical trials matched to a company, each tagged sponsorRole (lead vs collaborator; only lead trials are its own pipeline) |
| Tool | Description |
|---|---|
get_quotes_batch | Batch real-time quotes for up to 200 tickers (POST) |
get_quotes_universe | Quotes for an entire index universe (sp500/ndx/dji) |
get_index_snapshot | Bulk ticker snapshots (price, day change, % change) |
get_sector_etf_snapshot | Snapshot of the 11 SPDR sector ETFs with sector labels |
get_top_movers | Top gainers / losers / most-active by volume (with optional session window) |
get_market_breadth | Advance/decline, % above SMA50/200, 52-week high/low counts |
| Tool | Description |
|---|---|
get_earnings_calendar | Upcoming/recent earnings releases between two dates |
get_economic_calendar | Macro/economic events (CPI, FOMC, jobs, GDP) by date and country |
get_filing_calendar | 10-K / 10-Q SEC filing-deadline calendar by universe |
get_lockup_expirations | Upcoming IPO/secondary lockup expirations |
get_recent_material_filings | Recent material 8-K filings by index universe and item codes |
Backed by Signal8's own archive of SEC EDGAR filings — the archive's held range (see archiveFrom in each response; currently from late 2021, expanding back to 2016 as the history backfill fills in), full-text indexed and updated through the day. Filings older than archiveFrom may be absent, so an empty result is not proof a filing was never made. Record-only rows (bodyStored: false) carry metadata but no body.
| Tool | Description |
|---|---|
search_sec_filings | Filings by CIK, form, date, 8-K items, excluded forms, full-text q (match exact/stemmed, sort newest/relevance, facets counts by form/year/8-K item, understand on/off, section search, exhibitTypes — see query syntax below), and company filters (market cap, float, price, sector, industry, exchange, company type, foreign issuer, recent IPO) |
get_filing_documents | Every document in a filing (type, size, stored, secUrl) with the filingId / exhibitId to read each one |
get_filing_structured | Parsed Form 3/4/5 transactions, 13F holdings (values normalised to dollars), XML SCHEDULE 13D/G |
get_filing_document | Full text of a filing by filing id |
get_filing_exhibits | Readable exhibits of a filing, with ids |
get_exhibit_content | Full text of one exhibit |
search_filing_text | Exhibit-level full-text search returning the matching passage (match exact/stemmed, sort, section, exhibitTypes) |
lookup_accession_number | Accession number → filing id and metadata |
screen_sec_filings | Filings across companies by sector / industry / exchange / market cap |
screen_sec_filings_performance | Price returns +1/+3/+7/+30 days after filings (numbers, percent points), with aggregates |
Full-text query syntax (search_sec_filings.q, search_filing_text.pattern): unquoted words must all appear; "quoted phrase" for exact wording; OR between alternatives; -word to exclude; w1 NEAR/10 w2 (or NEAR(10), bare NEAR = 10) for proximity; word* for a prefix of 4+ letters; $AAPL or a capitalised ticker scopes to that company; a form code (8-K), item 5.02 / 8-K 1.05 and a bare year scope to that form / 8-K item / year. Relevance page 1 shows a few filings per company; meta.diversity names who has more. How the archive read the query is returned in meta.understanding (pass understand: false to search the words literally). facets: true adds per-form / per-year / per-8-K-item counts over the whole result set — absent when not measured, never zero.
Section search + exhibit types (both search tools): section restricts matching to business, risk_factors, cybersecurity, legal_proceedings, mdna, market_risk, controls, ex99 (press releases) or an 8-K item (item_5_02); exhibitTypes keeps filings with e.g. EX-10 (material contracts) or EX-99.1. A section is applied only when meta.sections says so — otherwise meta.unsupported explains that the whole filing was searched. Rows may carry matchedSections (absent = not measured).
Company filters on search_sec_filings follow the same rules as the signal8.ai/sec-filings page: they are applied from the archive only, against the filer's own CIK, on issuer-filed forms only (Form 4, 144 and SCHEDULE 13D/G are dropped and reported in companyFilter.droppedForms). Companies with no measured value for a filter are excluded and counted in companyFilter.unknown. Unknown values return a 400 listing the valid names. If the archive cannot answer, the call fails with COMPANY_FILTER_NOT_APPLIED and a reason rather than returning unfiltered rows.
13F values: the SEC changed the unit of a 13F from thousands of dollars to whole dollars for filings on/after 2023-01-03. get_filing_structured returns valueUsd / totalMarketValueUsd normalised to dollars (null when the unit cannot be determined) next to the raw valueThousands, and valueUnit says which unit the raw number is in.
| Tool | Description |
|---|---|
get_extractions | AI-extracted SEC filing data (13 types) |
get_filing_extractions | Extractions for a specific SEC filing |
get_extraction_dashboard | Cross-company extraction analytics |
get_extraction_by_type | Filter extractions by type (warrants, convertibles, etc.) |
Small-cap dilution analysis read directly from SEC filings. Every figure carries the accession number, filing date and the quoted sentence it came from. Included in Signal8 Pro. Call get_dilution_coverage first: it is the cheapest way to learn whether there is anything to fetch.
| Tool | Description |
|---|---|
get_dilution_coverage | Whether a ticker has a dilution analysis, how fresh it is, and why not if it doesn't |
get_dilution_risk | Dilution Pressure Score with its seven-component breakdown and risk levels. A partial score is out of scoreMaxMeasured (never rescale to /100); overheadSupply is total potential dilution excluding shelf (gross warrants/convertibles + ATM/equity line/S-1 at the current price, details.overheadSupplyBasis); levels named in levelsLowerBoundScope are floors ("at least") when levelsAreLowerBound is true; details.levelBounds gives each level's floor, ceiling and best estimate (a level whose floor and ceiling agree is exact) |
get_dilution_snapshot | Potential new shares and % of shares outstanding, cash position and runway, float |
get_dilution_instruments | One instrument family's rows with per-field SEC citations (warrants, convertibles, preferred, shelfs, atms, elocs, s1_offerings, equity_plans, exchangeables, recent_offerings) |
| get_baby_shelf_capacity | S-3 I.B.6 baby shelf: how much the issuer can raise off it
只显示了开头部分;其余内容请在仓库中阅读