Robinhood Chain activity alerts from RPC: density metrics, thresholds, cooldown, and dedupe

In this post, a “spot” means an observable, measurable object on Robinhood Chain—a pool contract, a watch-address balance, swap / Transfer logs in a post-launch time window, block-production rhythm—not a stock/coin pick, a “copy-trade position,” or a sniping playbook.

Goal: on top of Part 1 (newHeads heartbeat) and Part 2 (mint / pool-create logs), turn “high-volatility windows” into metrics + thresholds + cooldown alerts. Hearing a spike ≠ you should buy; alerts only guarantee that public facts got noisier.

Use	Endpoint
Public HTTPS	https://robinhood-mainnet-rpc.blockreq.com/v1/rpc/public
Public WSS	wss://robinhood-mainnet-rpc.blockreq.com/v1/rpc/public
Private HTTPS / WSS	Replace trailing public with your {API_KEY} (do not put it in demos)

Chain identity: EVM, chainId = 0x1237 (decimal 4663). Docs: Public endpoints · eth_subscribe · Supported Networks.

Series split

- Part 1: How to watch opens · RPC subscribe (robinhood-open-watch-rpc-subscribe)—newHeads + optional narrow logs. - Part 2: Listen for meme / new-token launches · logs and mint signals (robinhood-meme-mint-listen-logs)—topic filters, getLogs backfill. - This post (Part 3): turn those signals into metrics and alerts for monitorable spots (frequency, cooldown, dedupe). - Solana: Watch new Solana mints in the browser—another chain’s logsSubscribe; different protocol—don’t mix clients.

1. What a “spot” means (observable objects, not stock picks)

Treat a “spot” as a config item you are willing to measure continuously, not “an asset we recommend you hold”:

“Spot” dimension	On-chain observable	What you measure
Liquidity / pool	Pool or pair contract address; Swap / Sync / Mint / Burn logs (per real ABI)	Swaps per unit time, rough amount spikes, sudden silence
Watch account	Vault, MM, or self-custody wallet addresses	Native balance changes, ERC-20 Transfer in/out, nonce jumps
Post-launch window	N blocks or T minutes after a mint / PairCreated	Event density in the window, whether it crosses your “noisy” threshold
Tape rhythm	Height and timestamp deltas from newHeads	Whether block intervals stretch, whether gasUsed looks abnormal (congestion feel)

Recommended mindset:

• Monitorable spot = address / topics / time window in config + a metric formula.
• Not a stock pick = don’t name specific memes or imply “add size here.”
• High volatility = a metric steepens vs baseline; that is not opportunity, and risk is not priced.

Split “metric fired” from “should I trade”: this post only covers the first half—observation stack, not trading stack.

2. RPC / log metrics you can pull

On Robinhood Chain (EVM), metrics are almost always composed from these RPC families:

Metric family	Primary methods	Raw signals	Example derived measures
Block rhythm	WSS eth_subscribe → newHeads; or poll eth_blockNumber	number, timestamp, gasUsed	Δt (seconds between heads), blocks/min, gas utilization
Swap / event stream	WSS logs; HTTPS eth_getLogs	address + topic0 (Swap / Transfer…)	events/min, activity after txHash dedupe
Balance change	HTTPS eth_getBalance; optional eth_call for ERC-20 balanceOf	Hex wei	`
Account activity	Narrow logs related to that address; or eth_getTransactionCount	Transfer / custom events; nonce	In/out counts, nonce jumps
Backfill completeness	Chunked eth_getLogs	Gap block heights	Missed-sub duration, backfill count (ops metrics)

2.1 Balance snapshots (HTTPS, low frequency is fine)

curl -sS https://robinhood-mainnet-rpc.blockreq.com/v1/rpc/public \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"eth_getBalance","params":["0xYourWatchAddress","latest"]}'

Record returned result (hex wei) into a time series; alert on change magnitude / change frequency, not on a single absolute balance as a “buy/sell signal.”

2.2 Swap / Transfer density (WSS logs → sliding-window counts)

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "eth_subscribe",
  "params": [
    "logs",
    {
      "address": "0xYourPoolOrTokenAddress",
      "topics": ["0xYOUR_SWAP_OR_TRANSFER_TOPIC0"]
    }
  ]
}

• ERC-20 Transfer topic0 (standard): 0xddf252ad1be2c89b69c2b068fc378daa952ba7f163c4a11628f55a4df523b3ef
• DEX Swap / Sync, etc.: compute topic from that pool’s real ABI; write placeholders as 0xYOUR_SWAP_TOPIC0—don’t copy another chain’s factory address.

Common client-side derivation:

// Pseudocode: sliding-window event density
const windowMs = 60_000;
const timestamps = []; // push Date.now() on each push

function eventsPerMinute() {
  const now = Date.now();
  while (timestamps.length && now - timestamps[0] > windowMs) timestamps.shift();
  return timestamps.length; // ≈ events in last 1 minute (by arrival time; production may use block timestamp)
}

2.3 Block rhythm (wires into Part 1 newHeads)

// Pseudocode: seconds between adjacent newHeads
let prevTs = null;
function onHead(head) {
  const ts = parseInt(head.timestamp, 16);
  if (prevTs != null) {
    const gapSec = ts - prevTs;
    // sudden large gapSec → “blocks slowed” ops alert (not a trading signal)
  }
  prevTs = ts;
}

2.4 Metric-design notes

1. Narrow filters first — naked whole-chain Transfer will blow public quota in high vol and distort metrics.
2. Dedupe keys — txHash + logIndex (logs) or blockHash (heads); otherwise reconnect replay creates fake spikes.
3. Time anchor — demos can use local clock; serious dashboards prefer block.timestamp and accept skew vs wall clock.
4. Decode amounts carefully — data / indexed args need ABI decoding; a mis-decoded “huge swap” is worse than a missed alert. Examples here favor count-based metrics to avoid false precision.

3. Alert-threshold ideas (frequency, cooldown, dedupe)

The alert triad matters more than “what number is the threshold”:

Mechanism	Role	Typical default (sketch—tune yourself)
Frequency threshold	What counts as “noisy”	≥ N narrowly filtered logs in last 60s; or `
Cooldown	Don’t spam the same spot	After fire, suppress same alertKey for T seconds
Dedupe	Replay / multi-source	Set or LRU: txHash:logIndex; alert side also alertKey + bucket

Extra practical items:

• Hysteresis — fire on a high threshold, clear on a lower one, to avoid boundary chatter.
• Quiet baseline — collect p50/p95 on a dry run before picking N; density is naturally high in the first minutes after launch—give “time-window spots” their own thresholds.
• Channel degradation — on public 429, downsample and widen windows instead of retry-looping your IP into the ground.

3.1 Threshold state machine (pseudocode)

const COOLDOWN_MS = 120_000;
const THRESHOLD_EPM = 30; // events in last ~1 minute
const lastSent = new Map(); // alertKey -> lastSentAt

function maybeAlert(alertKey, epm, detail) {
  if (epm < THRESHOLD_EPM) return;
  const now = Date.now();
  const prev = lastSent.get(alertKey) || 0;
  if (now - prev < COOLDOWN_MS) return; // cooling down
  lastSent.set(alertKey, now);
  // send to Webhook / logs / desktop notification—log only here, no trading
  console.log(JSON.stringify({
    type: "high_vol_position",
    alertKey,
    epm,
    detail,
    ts: new Date(now).toISOString(),
    disclaimer: "observable_only_not_investment_advice",
  }));
}

3.2 Copyable mini-script: sliding window + cooldown (Node, public WSS)

Save as monitor-alerts.mjs, replace the placeholder with your watch contract address; connects only to public—no key. Needs Node 18+ (built-in fetch; see comment for WebSocket).

// monitor-alerts.mjs — observation alert demo; not investment advice, no orders
import WebSocket from "ws"; // Node 22+ can use global WebSocket instead

const WSS = "wss://robinhood-mainnet-rpc.blockreq.com/v1/rpc/public";
const ADDRESS = process.env.WATCH_ADDRESS || "0xYourPoolOrTokenAddress";
const TOPIC0 = process.env.TOPIC0 ||
  "0xddf252ad1be2c89b69c2b068fc378daa952ba7f163c4a11628f55a4df523b3ef";
const WINDOW_MS = 60_000;
const THRESHOLD = Number(process.env.THRESHOLD || 20);
const COOLDOWN_MS = Number(process.env.COOLDOWN_MS || 120_000);

if (!/^0x[a-fA-F0-9]{40}$/.test(ADDRESS) || ADDRESS.includes("YourPool")) {
  console.error("set WATCH_ADDRESS=0x… (real watch contract); refusing placeholder");
  process.exit(1);
}

const seen = new Set();
const stamps = [];
let lastAlertAt = 0;
let nextId = 1;
let backoff = 1000;

function trim(now) {
  while (stamps.length && now - stamps[0] > WINDOW_MS) stamps.shift();
}

function considerAlert(meta) {
  const now = Date.now();
  trim(now);
  const epm = stamps.length;
  if (epm < THRESHOLD) return;
  if (now - lastAlertAt < COOLDOWN_MS) return;
  lastAlertAt = now;
  console.log(JSON.stringify({
    alert: "high_event_density",
    address: ADDRESS,
    epm,
    threshold: THRESHOLD,
    cooldownMs: COOLDOWN_MS,
    meta,
    note: "observable spike; not investment advice",
  }));
}

function connect() {
  const ws = new WebSocket(WSS);
  ws.on("open", () => {
    backoff = 1000;
    const id = nextId++;
    ws.send(JSON.stringify({
      jsonrpc: "2.0",
      id,
      method: "eth_subscribe",
      params: ["logs", { address: ADDRESS, topics: [TOPIC0] }],
    }));
    console.error("subscribed logs", ADDRESS.slice(0, 10) + "…");
  });
  ws.on("message", (buf) => {
    let msg;
    try { msg = JSON.parse(String(buf)); } catch { return; }
    if (msg.method !== "eth_subscription") return;
    const r = msg.params && msg.params.result;
    if (!r || !r.transactionHash) return;
    const key = r.transactionHash + ":" + r.logIndex;
    if (seen.has(key)) return;
    seen.add(key);
    if (seen.size > 50000) seen.clear();
    const now = Date.now();
    stamps.push(now);
    trim(now);
    considerAlert({ tx: r.transactionHash, block: r.blockNumber, epm: stamps.length });
  });
  ws.on("close", () => {
    console.error("ws closed; reconnect in " + backoff + "ms");
    setTimeout(connect, backoff);
    backoff = Math.min(backoff * 2, 30000);
  });
  ws.on("error", () => { try { ws.close(); } catch (e) {} });
}

connect();

Run example:

# On Node 22+ with global WebSocket, you can drop the import WebSocket line
WATCH_ADDRESS=0xYourRealContract THRESHOLD=25 COOLDOWN_MS=180000 node monitor-alerts.mjs

Extensions (still observation stack):

• Also subscribe newHeads and fire ops alerts when gapSec > threshold (wires into Part 1).
• Give mint-from-zero (Part 2) its own alertKey so it doesn’t mix with plain Transfer density.
• Webhooks should only POST JSON logs; do not auto-trade inside the callback.

3.3 Minimal browser alert bar (optional)

Same single-page style as Parts 1–2: subscribe narrow logs, show “events in last 1 minute” locally, highlight when over threshold + cooldown. Save as robinhood-high-vol-monitor.html:

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8" />
  <title>Robinhood Chain · high-vol spot monitor (not investment advice)</title>
  <style>
    body { font-family: ui-sans-serif, system-ui, sans-serif; margin: 1.5rem; max-width: 52rem; }
    button { margin-right: 0.5rem; }
    #status { margin: 0.75rem 0; color: #334155; }
    #epm { font-size: 1.5rem; font-weight: 600; }
    .alert { color: #b91c1c; }
    pre { background: #0f172a; color: #e2e8f0; padding: 1rem; min-height: 12rem;
          overflow: auto; font-size: 12px; border-radius: 8px; }
    code { background: #f1f5f9; padding: 0.1rem 0.35rem; border-radius: 4px; }
  </style>
</head>
<body>
  <h1>Monitorable spot · event-density alerts</h1>
  <p>Public WSS · observe only · <strong>no stock picks, no orders, no return guarantee</strong></p>
  <p>WSS: <code>wss://robinhood-mainnet-rpc.blockreq.com/v1/rpc/public</code></p>
  <p><input id="addr" placeholder="0x… pool or token (required)" size="48" /></p>
  <p>Threshold <input id="th" type="number" value="20" min="1" style="width:4rem" /> / min · cooldown sec
    <input id="cd" type="number" value="120" min="1" style="width:4rem" /></p>
  <button id="start" type="button">Start</button>
  <button id="stop" type="button" disabled>Stop</button>
  <div id="status">Idle</div>
  <div>Events in last 1 min: <span id="epm">0</span></div>
  <pre id="out"></pre>
  <script>
    const WSS = "wss://robinhood-mainnet-rpc.blockreq.com/v1/rpc/public";
    const TRANSFER =
      "0xddf252ad1be2c89b69c2b068fc378daa952ba7f163c4a11628f55a4df523b3ef";
    const out = document.getElementById("out");
    const status = document.getElementById("status");
    const epmEl = document.getElementById("epm");
    let ws, want = false, nextId = 1, backoff = 1000, lastAlert = 0;
    const seen = new Set();
    const stamps = [];

    function log(line) {
      const t = new Date().toISOString().slice(11, 19);
      out.textContent = "[" + t + "] " + line + "\n" + out.textContent;
    }
    function trim(now) {
      while (stamps.length && now - stamps[0] > 60000) stamps.shift();
      epmEl.textContent = String(stamps.length);
    }
    function send(method, params) {
      ws.send(JSON.stringify({ jsonrpc: "2.0", id: nextId++, method, params }));
    }
    function connect() {
      if (!want) return;
      status.textContent = "connecting…";
      ws = new WebSocket(WSS);
      ws.onopen = () => {
        backoff = 1000;
        const addr = document.getElementById("addr").value.trim();
        if (!/^0x[a-fA-F0-9]{40}$/.test(addr)) {
          log("enter a valid address");
          want = false;
          document.getElementById("start").disabled = false;
          document.getElementById("stop").disabled = true;
          ws.close();
          return;
        }
        send("eth_subscribe", ["logs", { address: addr, topics: [TRANSFER] }]);
        status.textContent = "connected · observing (≠ investment advice)";
      };
      ws.onmessage = (ev) => {
        let msg; try { msg = JSON.parse(ev.data); } catch (e) { return; }
        if (msg.method !== "eth_subscription") return;
        const r = msg.params && msg.params.result;
        if (!r || !r.transactionHash) return;
        const key = r.transactionHash + ":" + r.logIndex;
        if (seen.has(key)) return;
        seen.add(key);
        const now = Date.now();
        stamps.push(now);
        trim(now);
        const th = Number(document.getElementById("th").value) || 20;
        const cd = (Number(document.getElementById("cd").value) || 120) * 1000;
        if (stamps.length >= th && now - lastAlert >= cd) {
          lastAlert = now;
          epmEl.classList.add("alert");
          log("ALERT epm=" + stamps.length + " th=" + th + " tx=" + r.transactionHash.slice(0, 12) + "…");
        }
      };
      ws.onclose = () => {
        status.textContent = "closed";
        if (!want) return;
        status.textContent = "reconnect in " + backoff + "ms";
        setTimeout(connect, backoff);
        backoff = Math.min(backoff * 2, 30000);
      };
      ws.onerror = () => { try { ws.close(); } catch (e) {} };
    }
    document.getElementById("start").onclick = () => {
      want = true;
      document.getElementById("start").disabled = true;
      document.getElementById("stop").disabled = false;
      connect();
    };
    document.getElementById("stop").onclick = () => {
      want = false;
      document.getElementById("start").disabled = false;
      document.getElementById("stop").disabled = true;
      epmEl.classList.remove("alert");
      if (ws) try { ws.close(); } catch (e) {}
      status.textContent = "stopped";
    };
  </script>
</body>
</html>

StackBlitz slot: TBD—the example is planned under blockreq/blog-demos at examples/robinhood-high-vol-monitor; for now, run the HTML / monitor-alerts.mjs above.

4. How to wire BlockReq query / subscribe (public + keyed)

Item	Public …/public	Private …/{API_KEY}
HTTPS queries	eth_getBalance, eth_blockNumber, eth_getLogs, eth_call	Same methods; higher plan limits
WSS subscribe	eth_subscribe: newHeads, logs	Same; fits 24×7 alerts
Rate limit	Best-effort, per IP	Per plan RU / req/s
Key	None	Create in Dashboard; never put it in public HTML / repos / screenshots
Fit	Learning, short high-vol window experiments	Long-running “monitorable spot” dashboards and alerts

Wiring tips (same as the series):

1. Subscribe for low latency; HTTP for gaps — after WSS disconnect, chunked eth_getLogs backfill, then recompute density (don’t only splice on local clock).
2. One narrow sub per spot — address + needed topics; multiple spots via a config list, not one “whole-chain logs” sub.
3. Separate alert process from UI — browser pages are fine for watching; cooldown alerts fit a small script / worker so sleep-dropped subs aren’t misread as “calm.”
4. Switch to keyed — change only the URL suffix; keep the key in env vars (e.g. replace public with your key placeholder).
5. 429 / instant disconnect — narrow width, raise cooldown and sampling interval, switch to keyed at peaks; don’t stack a dozen wide subs on one IP.

Backfill sketch (POST HTTPS public):

{
  "jsonrpc": "2.0",
  "id": 3,
  "method": "eth_getLogs",
  "params": [{
    "fromBlock": "0xFROM",
    "toBlock": "0xTO",
    "address": "0xYourPoolOrTokenAddress",
    "topics": ["0xYOUR_SWAP_OR_TRANSFER_TOPIC0"]
  }]
}

More: Public endpoints · Plans & pricing.

5. Risk disclaimer: no return guarantee, not investment advice

This post and sample code are for observable monitoring of public on-chain data and engineering demos only. They are not investment advice, stock/coin recommendations, or any promise of returns.

Please be clear:

1. No return guarantee — high event density, large balance moves, and slower blocks do not imply profitability; high volatility often comes with thin liquidity, fake books, contract risk, and information asymmetry.
2. Not investment advice — “spot” here = monitoring config, not a recommended holding; we do not provide or recommend any specific token or pool address.
3. No manipulation playbooks — no sniping, wash trading, volume washing, or manuals for bypassing Robinhood / DEX / platform risk controls.
4. Keys and compliance — don’t put API keys in public demos; whether trading is allowed and how you disclose it follows local law and platform rules.
5. Tool limits — public RPC is best-effort; log delay, replay, and ABI mis-decoding all produce false alerts or missed alerts.

Compliance boundary (same as the series): observable ≠ tradable signal; monitoring stack ≠ trading stack.

---

Summary

1. A “monitorable spot” = configurable observables (pool / watch address / post-launch window / block rhythm)—not a recommended position.
2. Metrics come from newHeads, narrow logs, eth_getBalance / eth_getLogs, etc.; prefer counts and intervals; amount decoding needs a real ABI.
3. Alerts rely on frequency threshold + cooldown + dedupe; sample scripts and HTML only log / highlight—no auto-trading.
4. BlockReq: short runs on public; long-running alerts on keyed; subscribe + getLogs backfill split; keys stay out of the repo.
5. Continue with: Part 1 open watch · Part 2 mint/logs · Solana new mint.

Typical next steps: spot list in config → baseline stats → Webhook alerts → persist checkpoints. Still an observation stack; no return guarantee, not investment advice.