SEC-AI Skill

40+ SEC filing tools. Access 10-K, 10-Q, 8-K, 13D/G beneficial ownership, insider transactions, company XBRL facts, and institutional holdings.

Github Open-Source

Please star Github if you like the skill.

Also available for OpenWebUI.

https://github.com/lkcair/sec-finance-ai

Setup

Run from the skill directory:

python3 -m venv .venv
.venv/bin/python3 -m pip install -r requirements.txt

> Windows: use .venv\Scripts\python3 instead of .venv/bin/python3.

AGNOSTIC OS / One-shot Design ( Goals )

• This skill is designed to be agnostic to OS and execution environment.
• It loads only essential context on first use and uses a single-shot interaction pattern for reliability.
• Basic usage pattern provided and All Functions list in SKILL.md body.

Quotation / Command Execution Reliability (common pitfall)

• If a command invocation uses complex shell quoting, it may fail in various environments. Use a here-doc style or a tiny helper script to avoid escaping issues.
• Crucially, ensure you are executing Python scripts using the interpreter from the skill's virtual environment. Simply calling python3 might not use the correct interpreter if the venv is not activated, leading to ModuleNotFoundError.
• Always use the full path to the venv's Python interpreter for execution, e.g.:

/home/openclaw/.openclaw/venv/sec-finance-ai/bin/python3

• Ensure that all necessary packages (like pandas, pydantic, requests, beautifulsoup4) are installed within this specific venv using its associated pip command.

Agent Quick-Start

After setup, copy the template below into your agent's TOOLS.md.

Replace SKILL_DIR with the absolute path to this skill's directory.

# SEC-AI Skill

## Usage

cd SKILL_DIR && SKILL_DIR/.venv/bin/python3 -c "

import asyncio, sys

sys.path.insert(0, '.')

from sec_finance_ai import Tools

t = Tools()

async def main():

result = await t.METHOD(ARGS)

print(result)

asyncio.run(main())

" 2>/dev/null

Replace METHOD(ARGS) with any call below.

## Common Calls

| Need | Method |
|---|---|
| Annual Report (US) | `get_latest_10k(ticker='AAPL')` |
| Quarterly Report (US) | `get_latest_10q(ticker='AAPL')` |
| Current Report | `get_recent_8k_filings(ticker='AAPL')` |
| Beneficial Ownership | `get_beneficial_ownership(ticker='GME')` |
| Insider Transactions | `get_insider_transactions(ticker='TSLA')` |
| Company Facts | `get_company_facts(ticker='AAPL')` |
| Filing Content | `get_filing_content(ticker='AAPL', get_full_content=True)` |
| Search Filings | `search_filings(query='Apple')` |
| Download Text | `download_filing_text(filing_url='...')` |

## Routing

- "Get 10-K" → `get_latest_10k`
- "Get 10-Q" → `get_latest_10q`
- "Get 8-K" → `get_recent_8k_filings`
- "Beneficial ownership" → `get_beneficial_ownership`
- "Insider trading" → `get_insider_transactions`
- "Financials/Facts" → `get_company_facts`
- "Search SEC" → `search_filings`
- "Download filing" → `download_filing_text`
- "Read full 10-K" → `get_filing_content(ticker='...', filing_type='10-K', get_full_content=True)`
- "Get links only" → `get_company_filings`

All Functions

By Categories:

• Filing Access: get_company_filings, get_latest_10k, get_latest_10q, get_recent_8k_filings, analyze_8k_filing, get_proxy_statements
• Ownership/Insider: get_insider_transactions, get_beneficial_ownership
• Financial Facts: get_company_facts, get_company_concept, get_available_metrics
• Content Retrieval: get_filing_content, download_filing_text
• Utility: search_filings, get_recent_ipos, get_sec_api_status, run_self_test, get_available_functions

Notes

• Data directly from official SEC EDGAR database.
• All functions are async — the asyncio.run() wrapper handles this.
• Remember to update SEC_HEADERS in sec_finance_ai.py with your User-Agent info.