Documentation

Master Every Feature

API Cookbook

Ready-made workflows for assistants, scripts and notebooks

Once Profitabul is connected (https://mcp.profitabul.com), these workflows appear in your assistant’s prompt menu. In Claude, type / and pick one. You only see the ones your plan supports.

  • pre_market_briefing: Context, key levels, today’s news and the edges that hold, on one page.
  • edge_hunt: Scan every report for a symbol and dig into the three strongest outcomes.
  • gex_game_plan: Gamma regime, walls and zero gamma, with scenarios, and alerts if you want them.
  • set_level_alerts: Suggests alerts at today’s key levels and creates the ones you approve.
  • strategy_stress_test: Backtest, walk-forward and Monte Carlo in the background, then a verdict.
  • journal_review: Your week in trades: what made money, what cost money, one change.

The Profitabul skill teaches Claude which tool fits which question, the rules that keep answers right (sample sizes, asking before changes, handling limits) and the workflows above. Add it in Claude under Settings → Capabilities → Skills, or drop it in your Claude Code skills folder.

Download SKILL.md
mkdir -p ~/.claude/skills/profitabul
curl -o ~/.claude/skills/profitabul/SKILL.md https://profitabul.com/skills/profitabul/SKILL.md

Pull 5-minute bars and the latest footprint bars into DataFrames for research.

import os, requests, pandas as pd

API = "https://compute.profitabul.ai/v1/tools"
H = {"Authorization": f"Bearer {os.environ['PROFITABUL_API_KEY']}"}

def tool(name, **args):
    r = requests.post(f"{API}/{name}", json=args, headers=H, timeout=120)
    body = r.json()
    if not body.get("ok"):
        raise RuntimeError(body["error"])
    return body["data"]

hist = tool("get_history", symbol="NQ", timeframe="5m", days=20, max_candles=5000)
bars = pd.DataFrame(hist["candles"]["data"], columns=hist["candles"]["columns"])
flow = tool("get_orderflow", symbol="NQ", timeframe="5m", lookback_minutes=390, max_bars=78)
print(bars.tail())
print(len(flow["bars"]), "orderflow bars")
Pro Tip
Field names come back exactly as the tool documents them; GET /v1/tools includes each tool’s JSON schema, and https://compute.profitabul.ai/openapi.json generates a typed client.

Read the live stream and print each quote and snapshot as it arrives.

const res = await fetch("https://compute.profitabul.ai/v1/stream/market?symbols=ES,NQ", {
  headers: { Authorization: `Bearer ${process.env.PROFITABUL_API_KEY}` },
});
const decoder = new TextDecoder();
let buffer = "";
for await (const chunk of res.body) {
  buffer += decoder.decode(chunk, { stream: true });
  let i;
  while ((i = buffer.indexOf("\n\n")) >= 0) {
    const block = buffer.slice(0, i);
    buffer = buffer.slice(i + 2);
    const event = /^event: (.*)$/m.exec(block)?.[1];
    const data = /^data: (.*)$/m.exec(block)?.[1];
    if (event === "quote") console.log(JSON.parse(data));
    if (event === "snapshot") console.log("CVD", JSON.parse(data).session_cvd);
  }
}

Read today’s walls and set alerts at them. Alerts are checked around the clock and sent to your devices. Run it from cron before the open.

call() { curl -s "https://compute.profitabul.ai/v1/tools/$1" -H "Authorization: Bearer $PROFITABUL_API_KEY" \
  -H "Content-Type: application/json" -d "$2"; }

GEX=$(call get_options_gex '{"symbol":"NQ"}')
CALL_WALL=$(echo "$GEX" | jq '.data.levels.ceiling')
PUT_WALL=$(echo "$GEX" | jq '.data.levels.floor')

call manage_alert "{\"action\":\"create\",\"symbol\":\"NQ\",\"price\":$CALL_WALL,\"condition\":\"crosses_above\",\"name\":\"Call wall\"}"
call manage_alert "{\"action\":\"create\",\"symbol\":\"NQ\",\"price\":$PUT_WALL,\"condition\":\"crosses_below\",\"name\":\"Put wall\"}"
Pro Tip
Prefer an alert that follows the wall instead of a fixed price? Create a GEX alert: {"action":"create","kind":"gex","symbol":"NQ","gexCondition":{"type":"spot_crosses_level","level":"callWall","direction":"above"}}

Submit a walk-forward test, poll until it finishes, and keep the result.

JOB=$(call submit_job '{"tool":"walk_forward","args":{"symbol":"ES","strategy_type":"breakout","total_days":365}}' | jq -r '.data.job.id')
while :; do
  OUT=$(call get_job "{\"jobId\":\"$JOB\"}")
  STATUS=$(echo "$OUT" | jq -r '.data.job.status')
  [ "$STATUS" = queued ] || [ "$STATUS" = running ] || break
  sleep "$(echo "$OUT" | jq -r '.data.job.retryAfterSec // 10')"
done
echo "$OUT" | jq '.data.job.result' > "walk-forward-$(date +%F).json"

Add an https endpoint in Settings → Integrations (or call manage_webhook) and pick the events: alert.fired (price and GEX alerts), condition.fired (agent conditions) and timer.ran (your timers, with the assistant’s answer). Each delivery is a JSON POST signed with your signing secret:

POST /your-endpoint
Profitabul-Event: alert.fired
Profitabul-Signature: t=1762959600,v1=5f2b...

{"id":"evt_...","type":"alert.fired","createdAt":"2026-11-12T15:00:00.000Z",
 "data":{"source":"price","symbol":"NQ","title":"Price Alert: NQ","message":"NQ crossed above 21500"}}

Verify it before acting on it (Node):

import crypto from "node:crypto";

function verify(rawBody, header, secret) {
  const { t, v1 } = Object.fromEntries(header.split(",").map((p) => p.split("=")));
  const expected = crypto.createHmac("sha256", secret).update(`${t}.${rawBody}`).digest("hex");
  const fresh = Math.abs(Date.now() / 1000 - Number(t)) < 300;
  return fresh && crypto.timingSafeEqual(Buffer.from(v1), Buffer.from(expected));
}
Pro Tip
Answer with any 2xx within 5 seconds. An endpoint that fails 25 times in a row is switched off; turn it back on from Settings. Use “Send test” to check your receiver.

Scan a symbol, then save the best outcome’s report setup so it is one click away on the Reports page.

TOP=$(call scan_edges '{"symbol":"GC","days":365,"limit":1}' | jq '.data.outcomes[0]')
call manage_saved_report "$(jq -n --argjson t "$TOP" \
  '{action:"create", name:("GC: " + $t.outcome), reportType:$t.reportId, symbol:"GC",
    parameters:($t.drillThrough.arguments | del(.reportType, .symbol))}')"

More in AI Assistants & API.