The options-flow data layer for AI agents

23 MCP tools over the GammaRips engine: the curated overnight pool, opportunity surfaces, a queryable outcome database, and the methodology itself. Your agent reasons to its own conclusions — there is no pick endpoint, on purpose.

Taste it right now — free, no card, no key

claude mcp add --transport http gammarips https://gammarips-mcp-406581297632.us-central1.run.app/mcp

The anonymous tier serves the pool preview, daily reports, methodology playbooks, and reference tools with zero setup. Ask your agent for a morning brief and see what comes back.

The full data layer — outcome history, opportunity surfaces, exit-rule simulation, all 23 tools — is $39/mo with a 7-day free trial. Your API key arrives by email shortly after you subscribe.

Works today with Claude Code, Cursor, and any MCP client that can send an Authorization header. Consumer claude.ai and ChatGPT connector UIs need OAuth — it's on the roadmap; the anonymous tier works everywhere now.

Quick Start

Endpoint:https://gammarips-mcp-406581297632.us-central1.run.app/mcp

Transport:Streamable HTTP (legacy SSE at /sse)

Auth:Authorization: Bearer <your API key>

Start with:get_enriched_signals · get_opportunity_surface · get_playbook

# Claude Code

claude mcp add --transport http gammarips \
  https://gammarips-mcp-406581297632.us-central1.run.app/mcp \
  --header "Authorization: Bearer YOUR_API_KEY"

# Python (fastmcp)

from fastmcp import Client
from fastmcp.client.transports import StreamableHttpTransport

transport = StreamableHttpTransport(
    "https://gammarips-mcp-406581297632.us-central1.run.app/mcp",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
)
async with Client(transport) as client:
    pool = await client.call_tool("get_enriched_signals", {})
    print(pool)

Built-in Prompts

The server ships MCP prompts — ready-made workflows your agent can run over the tools. None of them returns a pick; each ends in a decision surface you reason about.

morning_brief

Regime check → today's pool → historical context by delta bucket → a briefing of the most interesting candidates, with data caveats.

analyze_candidate

Deep-dive one name: enrichment, excursion history, realized labels of similar setups, and the honest risks.

run_your_own_tournament

The engine's bracket-tournament selection pattern, run by YOUR agent against YOUR objective, horizon, and risk tolerance.

23 Tools

Every tool is leakage-checked: nothing your agent reads contains information that wasn't knowable at the time it's dated. The outcome database behind these tools holds 3,000+ labeled contracts across 50+ scan days — every pool candidate since April 2026, growing every trading day. Full parameter schemas are self-describing over MCP.

Live pool

Today's curated candidates, structured for machine reasoning.

get_enriched_signals

The curated pool for a scan date — thesis, technicals, flow, and recommended contract per name. Served from the leakage-safe enriched view.

get_signal_detail

Deep dive on one ticker: full narrative enrichment (news, thesis, catalyst) plus point-in-time features and the recommended contract.

get_overnight_signals

The raw overnight scan across 5,230+ tickers, before curation — filter by direction, score, or ticker.

get_freemium_preview

A minimal public teaser of the top pool names — ticker, direction, score, headline. The taste, not the meal.

Research substrate

The deep data a human never browses — this is what you're paying for.

get_pool_features

Point-in-time feature vectors for the labeled candidate pool — every field knowable at selection time, nothing after.

get_opportunity_surface

Realized excursion surfaces per historical setup: how far each contract actually ran (peak) and drew down (trough), so your agent learns what was possible.

query_outcomes

Query the labeled outcome database across horizons, dates, tickers, and feature filters — the raw material for your agent's own research.

get_outcome_summary

Cohort-shaped aggregate outcomes (grouped by delta bucket, momentum, horizon…) with sample sizes attached.

estimate_exit_rule

Simulate a target/stop/horizon exit rule against the historical pool — test YOUR exit idea before your money meets it.

get_regime_context

Point-in-time volatility regime for a scan date: VIX vs VIX3M, SPY trend, and the engine's regime rail evaluated on those values.

Methodology

How the engine thinks — as playbooks your agent can execute.

list_playbooks

List the published methodology playbooks.

get_playbook

Fetch one playbook in markdown — including the bracket-tournament selection pattern your agent can run against your own objective.

Performance & receipts

The track record, queryable — including the unflattering parts.

get_signal_performance

Forward outcome tracking per enriched signal — signal-level results across the whole pool, not a curated highlight reel.

get_win_rate_summary

Aggregate underlying-direction statistics for the broad pool over a lookback window, with the caveats attached.

get_position_history

Closed trades from the paper-trading validation cohort's ledger.

get_historical_performance

Ledger aggregates by window, direction, and policy version.

Reports & metadata

Context and contracts.

get_daily_report

The full daily intelligence report (markdown) — the editorial synthesis of the scan.

get_report_list

List available daily reports.

get_available_dates

Which scan dates have data.

get_enriched_signal_schema

The machine-readable data contract: every substrate column with its leakage classification and as-of boundary.

Reference & external

Utilities that kill hallucination classes.

get_market_calendar_status

Is the US market open today? Deterministic NYSE calendar — holidays and early closes included.

get_signal_explainer

Plain-English definition of any GammaRips signal field.

web_search

Google web search for real-time fact verification and grounding.

Pricing

FREE

The Website

✓ Today's curated pool, human-readable
✓ Daily reports + per-ticker deep dives
✓ Public scorecard + the Lab
✓ Methodology + full disclosures
✓ Free forever — it's not a trial

THE PRODUCT

Agent Access — MCP

$39/mo

✓ All 23 tools + the built-in prompts
✓ Opportunity surfaces + outcome database
✓ Exit-rule simulation + regime context
✓ Methodology playbooks
✓ 7-day free trial · cancel anytime

Give your agent something real to reason over

A chatbot with no data improvises. An agent with the pool, the surfaces, and the methodology does analysis. Connect yours in minutes.