TickerRisk MCP Server

Check an options trade for hidden catalysts before you sell premium — from inside Claude, ChatGPT, or Cursor.

An MCP server that lets an AI assistant answer questions like "is it safe to sell a 30-day put on HPE?" with real event data instead of a guess.

The problem it solves: a fat option premium is usually the market pricing in a known upcoming event — an earnings report, an FDA decision, a court date — not free money. Some screeners flag earnings; few check the wider event calendar. This one scores every candidate across earnings, FDA, legal, SEC and clinical events landing inside the expiry window, and filters out the traps.

You:    Is it safe to sell a 4-week put on INTC?

Claude: [calls scan_ticker]
        INTC scores 100/100 (HIGH) over a 4-week window. Earnings land in 4 days,
        inside your expiry. The premium is pricing that gap — 279% annualized on
        the $88 put is compensation for event risk, not an edge.

Why this is different

Risk is horizon-dependent, and that is the whole point. The same stock:

Ticker	1-week window	4-week window	Why it changes
AAPL	51 (MEDIUM)	97 (HIGH)	Earnings sit in week 3
KO	22 (LOW)	62 (MEDIUM)	Earnings enter the window

Sell a weekly and you are fine. Sell a monthly on the same ticker and you have sold straight through an earnings report. A screener that shows one number cannot tell you that.

Install

Claude Desktop

Add to your claude_desktop_config.json:

macOS/Linux — ~/Library/Application Support/Claude/claude_desktop_config.json Windows — %APPDATA%\Claude\claude_desktop_config.json

{
  "mcpServers": {
    "tickerrisk": {
      "command": "uvx",
      "args": ["tickerrisk-mcp"]
    }
  }
}

Restart Claude Desktop. You should see the TickerRisk tools in the tool menu.

Claude Code

claude mcp add tickerrisk -- uvx tickerrisk-mcp

Cursor

In ~/.cursor/mcp.json:

{
  "mcpServers": {
    "tickerrisk": { "command": "uvx", "args": ["tickerrisk-mcp"] }
  }
}

From source

git clone https://github.com/Tickerrisk/tickerrisk-mcp
cd tickerrisk-mcp
pip install -e .
tickerrisk-mcp

Tools

Tool	What it answers
scan_ticker	"How risky is selling an option on X over the next N weeks?"
find_wheel_candidates	"What cash-secured puts can I sell this week without an earnings trap?"
find_covered_calls	"What calls can I sell against shares I already own?"
compare_tickers	"Which of these stocks is safest to sell premium on right now?"

scan_ticker(ticker, expiry_weeks=4)

Returns a 0–100 catalyst-risk score (higher = riskier) with the events driving it: earnings date and whether it falls in the window, FDA/clinical milestones, legal filings, SEC events, implied volatility, IV Rank, and the expected move.

Bands: ≥70 HIGH · 45–69 MEDIUM · <45 LOW

find_wheel_candidates(week, risk, max_risk, min_put_oi, sector, limit)

Scans the S&P 500 for cash-secured puts and returns only names whose catalyst score over the option's own expiry window is under max_risk. Flags any candidate whose earnings land before expiry.

find_covered_calls(week, risk, max_risk, min_call_oi, sector, limit)

Same gating for the call side. Income is computed from time value only, so in-the-money strikes do not show inflated yields.

compare_tickers(tickers, expiry_weeks=4)

Side-by-side catalyst risk for up to 25 symbols on one horizon.

Access and authentication

No signup needed to start. Access follows tickerrisk.com's normal model:

• First 24 hours — full access, no account, keyed to your IP
• After that — a free account at tickerrisk.com adds 14 days
• The strategy scanners (find_wheel_candidates, find_covered_calls, compare_tickers) read cached data and stay available

To authenticate an existing account, set a token:

{
  "mcpServers": {
    "tickerrisk": {
      "command": "uvx",
      "args": ["tickerrisk-mcp"],
      "env": { "TICKERRISK_TOKEN": "your-jwt-here" }
    }
  }
}

Environment variables

Variable	Default	Purpose
TICKERRISK_TOKEN	(none)	JWT for an existing account
TICKERRISK_BASE_URL	https://tickerrisk.com	Override the API host
TICKERRISK_TIMEOUT	45	Request timeout in seconds

Data and limitations

Being straight about what this is and is not:

• Catalyst data — earnings dates, SEC filings, court records (CourtListener), ClinicalTrials.gov, and news. Public sources, so incomplete or delayed entries happen.
• Option premiums are indicative, not live NBBO. Quotes are roughly 15 minutes delayed and estimated where no bid exists (marked in the output). Always confirm in your broker before trading.
• Outside US market hours (9:30–16:00 ET) quotes go stale and candidate lists thin out. This is upstream data reality, not a bug.
• The score is not a prediction. It measures scheduled event exposure and volatility. A LOW score does not mean a stock cannot drop; it means no known catalyst was found in that window.

Not financial advice. For research only.

How it compares

Being accurate about this, because the differentiator is narrower than most tools claim:

Earnings-date checking is not unique. Barchart's options screener has a "Flag Earnings" option that marks contracts whose next earnings date falls on or before expiration. Market Chameleon tracks biotech catalysts and links them to option chains. If earnings alone is what you need, those are mature tools with real-time data and far more filters — use them.

What this tool does differently is combine five event types — earnings, FDA decisions, legal filings, SEC events and clinical milestones — into a single 0–100 score tied to your expiry window, and filter on it by default rather than showing an optional flag column. Court records as an options-risk input in particular is something we have not found elsewhere.

So: if you want the deepest screener, use Barchart or Option Samurai. If you want one number that answers "is there anything scheduled inside this expiry", that is what this is.

Links

• Website — tickerrisk.com
• Wheel scanner — tickerrisk.com/wheel
• Covered calls — tickerrisk.com/covered-calls

License

MIT

---

<!-- Ownership marker required by the official MCP Registry: it verifies control of the PyPI package by checking this exact string appears in the published README. Must match the "name" field in server.json. Do not remove or reword. -->

mcp-name: io.github.PasiutusVovere/tickerrisk-mcp