sec-edgar-mcp

10 years of XBRL financials on any public company in 6 seconds.

---

What Is This?

The SEC's EDGAR system contains more than 20 million filings from every public company, investment fund, and regulated entity that has ever filed with the Commission — going back to the early 1990s. It's the most comprehensive public financial database in existence. The problem is the interface: EDGAR's search is built for lookup, not analysis. Extracting structured financial data from a 10-K means parsing XBRL, navigating inline HTML, handling company name variations, and reconciling CIK numbers — before you've done any actual analysis.

This actor provides four modes of structured access to EDGAR. Filing search lets you full-text search across any filing type (10-K, 10-Q, 8-K, S-1, DEF14A, and 30+ others) with filters by company name, CIK, date range, and form type — returns document-level metadata ready to feed a downstream processing pipeline. Company financials pulls structured XBRL data — revenue, net income, total assets, EPS, and 15+ other metrics — going back up to 10 fiscal years, normalized and ready to analyze without touching a single XML file. Insider trading tracker pulls Form 4 filings by company or individual insider, giving you every reported open-market purchase and sale with transaction date, price, and shares. Recent filings monitor returns a live feed of all new filings matching a company or form type filter.

It runs in Batch mode for one-shot research pipelines and MCP server mode for live financial research in Claude Desktop or any MCP-compatible AI assistant. Financial analysts use it to build comparable company analyses in minutes. M&A teams use it to pull financials on acquisition targets without Bloomberg. Investigative journalists use it to track insider activity and follow the money through SEC disclosures.

---

Who Uses This

Buy-Side Financial Analysts You need to build a 5-year financial model for a mid-cap software company. Pulling revenue, gross profit, operating income, and EPS from 20 quarterly 10-Qs manually takes 4–6 hours. The financials mode returns all of it, structured and labeled, in under 10 seconds. You spend your time on the model, not the data extraction.

M&A Due Diligence Teams You're evaluating a public company as an acquisition target. You need the last 8 years of revenue and margin trends, every 8-K that mentions material contracts or litigation, and the last 12 months of insider selling activity. This actor handles all three data pulls in three calls — financials mode, filing search mode, and insider tracker mode — giving you a structured due diligence snapshot before you've even signed an NDA.

Hedge Fund Quantitative Researchers You're building a factor model that ingests structured financial data from 500+ public companies. EDGAR's XBRL is the cleanest source for normalized financials across filers, but parsing it at scale requires significant infrastructure. This actor exposes XBRL financials as a clean API call — batch across your universe, get structured output, feed your model directly.

Investigative Journalists and Financial Reporters You're investigating whether insiders at a pharmaceutical company sold shares before a failed drug trial announcement. The insider tracker mode returns every Form 4 filing for the company's officers and directors, with transaction dates, prices, and relationship to the filer — exactly what you need to build a timeline of unusual activity.

Venture Investors and Startup Analysts You need to understand the financial trajectory of every public company in a target sector before making a private market investment. The financials mode gives you 10-year revenue and margin trends for any public comp in seconds — building sector benchmarks that used to require a Bloomberg terminal.

---

Key Features

Feature	Detail
Mode 1: Filing search	Full-text search across 20M+ EDGAR filings by form type, company, CIK, date
Mode 2: Company financials	Structured XBRL data — 15+ metrics, up to 10 fiscal years, any public company
Mode 3: Insider trading	Form 4 data — every reported transaction by officers, directors, and 10%+ holders
Mode 4: Recent filings	Live feed of new filings by company or form type
Financial metrics	Revenue, net income, gross profit, operating income, EPS, total assets, total liabilities, stockholders' equity, cash and equivalents, operating cash flow, capex, and more
Form types supported	10-K, 10-Q, 8-K, S-1, S-1/A, DEF14A, 4, 13F, SC 13G, SC 13D, 20-F, 6-K, and 30+ others
Company lookup	By ticker, company name (partial match), or CIK number
Historical depth	Financial data to 2014 (10-year XBRL); filing metadata to 1994
MCP server mode	Live tool-use in Claude Desktop, Claude Code, any MCP client
Batch mode	One-shot runs for pipeline ingestion

---

Quick Start

Batch Mode — Company Financials

Input:

{
  "mode": "batch",
  "query_mode": "financials",
  "company": "Salesforce",
  "metrics": ["revenue_usd", "net_income_usd", "gross_profit_usd", "operating_income_usd", "eps_diluted"],
  "years": 8
}

Output:

{
  "entity_name": "Salesforce, Inc.",
  "cik": "1108524",
  "ticker": "CRM",
  "financials": [
    {
      "fiscal_year": 2024,
      "period_end": "2024-01-31",
      "revenue_usd": 34857000000,
      "net_income_usd": 4136000000,
      "gross_profit_usd": 26323000000,
      "operating_income_usd": 1731000000,
      "eps_diluted": 4.20,
      "form_source": "10-K",
      "filed_date": "2024-03-07"
    },
    {
      "fiscal_year": 2023,
      "period_end": "2023-01-31",
      "revenue_usd": 26492000000,
      "net_income_usd": 208000000,
      "gross_profit_usd": 19985000000,
      "operating_income_usd": -64000000,
      "eps_diluted": 0.21,
      "form_source": "10-K",
      "filed_date": "2023-03-09"
    }
  ]
}

Batch Mode — Insider Trading

Input:

{
  "mode": "batch",
  "query_mode": "insider_trading",
  "company": "NVDA",
  "transaction_types": ["P", "S"],
  "date_from": "2024-01-01",
  "max_results": 50
}

Output (truncated):

[
  {
    "entity_name": "NVIDIA Corporation",
    "cik": "1045810",
    "insider_name": "Huang Jen-Hsun",
    "insider_title": "President and Chief Executive Officer",
    "transaction_date": "2024-06-13",
    "transaction_type": "S",
    "shares": 120000,
    "price_per_share": 135.18,
    "total_value": 16221600,
    "shares_owned_after": 75982394,
    "form_type": "4",
    "file_date": "2024-06-17",
    "filing_url": "https://www.sec.gov/Archives/edgar/data/1045810/000104581024000049/0001045810-24-000049-index.htm"
  }
]

MCP Server Mode

{
  "mcpServers": {
    "sec-edgar": {
      "command": "npx",
      "args": [
        "apify-actor-mcp",
        "--actor-id", "your-username/sec-edgar-mcp",
        "--token", "YOUR_APIFY_TOKEN"
      ]
    }
  }
}

Then in Claude: "Pull the last 5 years of revenue and net income for Microsoft and compare their operating margin trend."

---

MCP Tools Exposed

search_filings

Full-text search across EDGAR filings by company, form type, date, and keyword.

{
  "tool": "search_filings",
  "arguments": {
    "company": "Palantir",
    "form_types": ["10-K", "10-Q"],
    "date_from": "2023-01-01",
    "date_to": "2024-12-31",
    "max_results": 10
  }
}

get_company_financials

Return structured XBRL financial data for any public company.

{
  "tool": "get_company_financials",
  "arguments": {
    "company": "AAPL",
    "metrics": ["revenue_usd", "net_income_usd", "total_assets_usd", "operating_cash_flow_usd"],
    "years": 10
  }
}

get_insider_transactions

Return Form 4 insider trading data for a company or specific insider.

{
  "tool": "get_insider_transactions",
  "arguments": {
    "company": "META",
    "transaction_types": ["P", "S"],
    "date_from": "2024-01-01",
    "max_results": 100
  }
}

monitor_recent_filings

Return the most recent filings for a company or set of form types.

{
  "tool": "monitor_recent_filings",
  "arguments": {
    "company": "Tesla",
    "form_types": ["8-K"],
    "max_results": 20
  }
}

get_filing_document

Retrieve the full text and metadata for a specific filing by accession number.

{
  "tool": "get_filing_document",
  "arguments": {
    "accession_number": "0001045810-24-000049"
  }
}

---

Input Schema

Field	Type	Default	Description
mode	string	"batch"	"batch" or "mcp"
query_mode	string	required	"search", "financials", "insider_trading", "recent_filings"
company	string	—	Company name, ticker, or CIK number
form_types	array	all	Filing form types: ["10-K"], ["10-Q","8-K"], etc.
date_from	string	—	ISO 8601 filing date start
date_to	string	—	ISO 8601 filing date end
keyword	string	—	Full-text keyword for filing search mode
metrics	array	all	Financial metrics for financials mode (see supported list below)
years	integer	5	Number of fiscal years for financials mode (max 10)
transaction_types	array	all	Insider transaction types: "P" (purchase), "S" (sale), "A" (award)
insider_name	string	—	Filter insider transactions by specific individual
max_results	integer	100	Maximum records to return
sort_by	string	mode-dependent	Sort field (varies by mode)

Supported financial metrics: revenue_usd, net_income_usd, gross_profit_usd, operating_income_usd, ebitda_usd, eps_basic, eps_diluted, total_assets_usd, total_liabilities_usd, stockholders_equity_usd, cash_and_equivalents_usd, operating_cash_flow_usd, capex_usd, long_term_debt_usd, shares_outstanding

---

Use Case Recipes

Recipe 1: Comparable Company Analysis (Comps)

Pull 5-year revenue and margin data for 5 SaaS comparables:

{
  "query_mode": "financials",
  "company": "Workday",
  "metrics": ["revenue_usd", "gross_profit_usd", "operating_income_usd", "net_income_usd"],
  "years": 5
}

Run for each comp (Salesforce, ServiceNow, Veeva, HubSpot) — output feeds directly into a comps table without manual data entry.

Recipe 2: Pre-Announcement Insider Activity

Check for unusual insider selling before a major announcement:

{
  "query_mode": "insider_trading",
  "company": "MRNA",
  "transaction_types": ["S"],
  "date_from": "2024-01-01",
  "date_to": "2024-06-30",
  "max_results": 200
}

Outcome: Complete insider sale record with prices, dates, and post-transaction ownership — build a timeline against any 8-K announcements.

Recipe 3: Material Event Monitoring

Track all 8-K filings for a portfolio of 10 companies:

{
  "query_mode": "recent_filings",
  "company": "TSLA",
  "form_types": ["8-K"],
  "max_results": 30
}

Outcome: Recent material event disclosures — run across your portfolio and pipe to a Slack alert or portfolio management system.

Recipe 4: S-1 Pipeline Monitoring

Monitor all new S-1 and S-1/A filings for upcoming IPOs in tech:

{
  "query_mode": "search",
  "form_types": ["S-1", "S-1/A"],
  "keyword": "software artificial intelligence",
  "date_from": "2024-01-01",
  "max_results": 50
}

Outcome: AI/software IPO pipeline with filing dates and company names — feed a deal sourcing workflow.

---

Connecting to Claude Desktop / Claude Code

Claude Desktop:

{
  "mcpServers": {
    "sec-edgar": {
      "command": "npx",
      "args": [
        "apify-actor-mcp",
        "--actor-id", "your-username/sec-edgar-mcp",
        "--token", "apify_api_YOURTOKEN"
      ],
      "env": {}
    }
  }
}

Claude Code:

{
  "mcpServers": {
    "sec-edgar": {
      "command": "npx",
      "args": ["apify-actor-mcp", "--actor-id", "your-username/sec-edgar-mcp", "--token", "apify_api_YOURTOKEN"]
    }
  }
}

Sample prompts once connected:

• "Compare the revenue growth rate for Apple vs. Microsoft over the last 8 years."
• "Show me every insider purchase at Nvidia in the last 6 months."
• "What 8-K filings has Tesla made in 2024?"
• "Pull the 10-year financial history for Amazon including operating cash flow."

---

Pricing

Volume	Price
First 1,000 results	$0.50
1,001 – 50,000 results	$0.40/1k
50,000+ results	$0.30/1k

ROI context: Bloomberg Terminal costs $24,000/year per seat. FactSet and S&P Capital IQ run $10,000–$30,000/year. This actor provides the same underlying EDGAR data — normalized, structured, and API-accessible — at a fraction of the cost. For a research team running 500 company lookups per month, total cost is under $25.

---

Data Source and Freshness

• Primary: SEC EDGAR full-text search system (efts.sec.gov) and EDGAR company facts API (data.sec.gov)
• XBRL source: SEC EDGAR company facts endpoint — standardized structured financial data reported directly by filers
• Update frequency: EDGAR updates continuously as filings are accepted; actor reflects filings within 1 hour of SEC acceptance
• Financial data depth: XBRL mandatory since 2009 for large accelerated filers; actor provides data back to FY2014 with high reliability
• Insider trading: Form 4 filings required within 2 business days of transaction; actor reflects same-day filing data

---

Technical Notes

• Company lookup supports ticker, full name, and CIK number — all are resolved to a canonical CIK internally
• XBRL data quality varies by filer; small companies and foreign private issuers (20-F) may have incomplete XBRL tagging
• Financial figures are in USD (or converted from functional currency where reported); always check the form_source field to know which filing the data came from
• EPS figures use diluted weighted average shares unless eps_basic is explicitly requested
• Insider transaction type codes: P = open-market purchase, S = open-market sale, A = award/grant, D = disposition to issuer, F = tax withholding
• Accession numbers use the format XXXXXXXXXX-YY-NNNNNN (e.g., 0001045810-24-000049)
• The SEC imposes rate limits of 10 requests/second; the actor manages throttling automatically

---

Support

• Issues: File a GitHub issue on the actor repository
• Apify Community: community.apify.com
• SEC EDGAR full-text search: efts.sec.gov/LATEST/search-index
• SEC EDGAR company facts API: data.sec.gov/api/xbrl/companyfacts
• EDGAR filing viewer: www.sec.gov/cgi-bin/browse-edgar