Use OOTWOracle with AI — MCP Server, API, ChatGPT, Claude, n8n

Integration Methods

🤖

MCP Server

Add to Claude Desktop, Cursor, or Windsurf. Query predictions directly in conversation.

⚡

ChatGPT GPT Actions

Import our OpenAPI schema into a custom GPT. Instant psychedelic intelligence assistant.

🔗

REST API

Plain JSON endpoints. No auth. Works with any HTTP client, LangChain, or curl.

🔄

n8n / Make.com

HTTP Request node → parse predictions array. Build daily briefing automations in minutes.

🦜

LangChain / LlamaIndex

Use requests tool or JsonLoader with our API. Drop into any RAG pipeline.

📡

RSS / JSON Feed

Subscribe to our feed in any RSS reader or feed aggregator. Auto-syncs to Medium.

API Endpoints — No Auth Required

All endpoints are public, free, and updated daily by noon ET.

GET

https://www.ootworacle.com/predictions.json

Today's 5 predictions with confidence scores, reasoning, and resolve dates.

GET

https://www.ootworacle.com/all_predictions.json

Full archive — 162+ predictions across all dates. Filter by category, date, or compound.

GET

https://www.ootworacle.com/report/{date}

Full HTML report for any date (e.g. /report/2026-05-27). 30 reports available.

GET

https://www.ootworacle.com/feed.json

JSON Feed of latest 20 oracle reports. Compatible with any feed reader.

GET

https://www.ootworacle.com/openapi.json

Full OpenAPI 3.0 schema. Import directly into ChatGPT GPT Actions or Composio.

GET

https://www.ootworacle.com/llms-full.txt

14 full oracle reports formatted for LLM ingestion. Ideal for RAG context.

MCP Server — Claude Desktop & Cursor

The OOTWOracle MCP server gives Claude and other MCP-enabled tools 5 callable tools: get today's predictions, fetch any report, search across all 162 predictions, check oracle status, and read the full agent debate.

Quick Install (2 minutes)

Install dependencies

pip install mcp httpx

Download the server

curl -o ootworacle_mcp.py https://www.ootworacle.com/mcp-server/server.py

Add to Claude Desktop config

Edit ~/.config/claude/claude_desktop_config.json (macOS: ~/Library/Application Support/Claude/claude_desktop_config.json)

{
  "mcpServers": {
    "ootworacle": {
      "command": "python",
      "args": ["/absolute/path/to/ootworacle_mcp.py"]
    }
  }
}

Restart Claude Desktop

You'll see OOTWOracle in your tools list. Ask Claude: "What does OOTWOracle predict about ibogaine this week?"

Available MCP Tools

get_todays_predictions()

Returns today's 5 predictions with confidence scores, reasoning, categories, and resolve dates.

get_oracle_report(date="YYYY-MM-DD")

Fetches the full intelligence report for any date. 30 reports available from 2026-04-28.

search_predictions(query="ibogaine", limit=10)

Searches all 162+ predictions by keyword. Find everything the Oracle has said about any compound or topic.

get_oracle_status()

Live status — today's headline, top predictions, and what the swarm is tracking right now.

ChatGPT GPT Actions

Go to chat.openai.com/gpts/editor

Create a new GPT or edit an existing one.

Add Action → Import from URL

https://www.ootworacle.com/openapi.json

Set system prompt (optional)

You are an oracle assistant with access to OOTWOracle — 
a daily AI prediction service on psychedelic medicine.
Use the get_predictions action to answer questions about
FDA timelines, clinical trials, biotech markets, and
drug policy. Always cite confidence scores.

n8n / Make.com / Zapier

n8n — Daily Briefing Automation

// HTTP Request node
Method: GET
URL: https://www.ootworacle.com/predictions.json

// Parse response
{{ $json.predictions[0].title }} — {{ $json.predictions[0].confidence }}%
{{ $json.predictions[0].reasoning }}

Make.com (Integromat)

Module: HTTP → Make a request
URL: https://www.ootworacle.com/predictions.json
Method: GET
Parse response: Yes
→ Map predictions[] array to your workflow

LangChain / Python

import requests

# Today's predictions
data = requests.get("https://www.ootworacle.com/predictions.json").json()
for p in data["predictions"]:
    print(f"{p['confidence']}% — {p['title']}")

# Search all predictions
archive = requests.get("https://www.ootworacle.com/all_predictions.json").json()
ibogaine = [p for p in archive["predictions"]
            if "ibogaine" in p["title"].lower()]

# Full report as LLM context
report = requests.get("https://www.ootworacle.com/llms-full.txt").text

What the Oracle Covers

OOTWOracle's 8-agent swarm tracks everything in the psychedelic medicine ecosystem:

🏛️ FDA & Regulation

Approval timelines, REMS updates, NDA filings, clinical holds, breakthrough designations

🧬 Research

Phase 1/2/3 trials, MAPS MDMA therapy, Compass psilocybin, ibogaine/oxa-noribogaine

📜 Legislation

Congressional bills, state right-to-try frameworks, VA reform, scheduling petitions

📈 Markets

CMPS, ATAI, MMED, MindMed valuations, biotech funding rounds, M&A signals

🔬 Compounds

Psilocybin, MDMA, ibogaine, ketamine, LSD, DMT — mechanism, safety, access

🌍 Culture

Veteran advocacy, harm reduction, patient communities, media narrative shifts

Ready to integrate?

Free API — no auth, no rate limits, no signup. Updated daily by 11:11am ET.

Try the API OpenAPI Schema MCP Server

Questions?

Contact: alex@ootw.co

More discovery files: /llms.txt · /.well-known/mcp.json · /agent-manifest.json · /.well-known/ai-plugin.json