🛡 StableHub
Check Token →

Changelog

Last updated: 11 June 2026

This is the public changelog for StableHub — every meaningful change to our tools, methodology, and infrastructure. New entries land at the top, dated, with a version where one applies. We publish this openly so users, journalists, and AI assistants can verify what has actually shipped versus what is on the roadmap. The Rug Risk Score methodology is the canonical reference for how scoring works in each release.

Recent updates

Rug Risk Score v1.1 — June 2026

  • Token age now an active signal. Sourced from a native-L1 launch-date map (21 chains including BTC, ETH, SOL), Etherscan’s getcontractcreation for ERC-20 contracts, and CoinGecko’s genesis_date as a final fallback. The “under 30 days → MEDIUM minimum” rule now fires reliably.
  • Top-10 holder concentration now an active signal, sourced from Moralis (/erc20/{address}/owners). v1.1 reports the raw percentage and exposes it as a flag in the verdict layer.
  • Caveat on top-10 concentration: the raw figure includes labelled exchange-custody wallets (Binance, Coinbase, Tether Treasury, Bitfinex MultiSig) and inflates for highly-traded stablecoins — USDT comes back ~49% with roughly 30 points of that being exchange custody. A custody-aware “non-custodial concentration” metric is on the v1.2 roadmap.
  • ~8 of the 50 currently-tracked tokens still return null for age (BSC-only / Solana-only / RWA listings) and now display “Data unavailable” in the risk table rather than silently scoring around the missing field.
  • Published methodology updated: stablehub.io/methodology/rug-risk-score.

Rug Risk Score v1.0 — June 2026

  • Initial launch of the StableHub Rug Risk Score engine.
  • Coverage: Ethereum mainnet only, ERC-20 tokens.
  • Data sources: CoinGecko (/coins/ethereum/contract/{address}) for market and metadata, Etherscan V2 (getsourcecode) for contract verification status.
  • AI verdict layer: Mistral Small 3.1 (mistral-small-latest) running in native JSON mode for structured output.
  • Output schema: riskLevel (LOW / MEDIUM / HIGH), the top three contributing flags, a plain-English verdict, and a disclaimer line attached to every response.
  • Validation set: 10/10 correctly classified across five known rug pulls (SafeMoon v1, SQUID, AnubisDAO, TurtleDex, Luna Yield) and five large-cap legitimate tokens (USDT, USDC, LINK, UNI, PEPE).
  • Published methodology: stablehub.io/methodology/rug-risk-score.
  • Live tool: stablehub.io/check/token.

Token checker tool — June 2026

  • Free contract address scanner now public at /check/token.
  • Typical response time: 2–4 seconds end-to-end.
  • Privacy: checked contract addresses are not logged or stored anywhere on our infrastructure.
  • Mobile-first interface — the primary use case is a quick check before buying, often on a phone.

Infrastructure — June 2026

  • Self-hosted n8n workflow engine deployed at n8n.stablehub.io, powering the Rug Risk Score pipeline and queued for future tools.
  • Public uptime monitor live at status.stablehub.io, running Uptime Kuma.
  • Auto-recovery on the rug-check webhook: a 15-minute health check automatically re-registers the endpoint if it stops responding, so a quiet failure does not become a long outage.

Site foundations — May 2026 (pre-launch)

  • Custom cookie consent banner deployed, covering UK GDPR, EU GDPR, CCPA/CPRA (California), PIPEDA (Canada), APP (Australia), and LGPD (Brazil), wired into Google Consent Mode v2.
  • Custom SMTP mailer plugin built for transactional email, removing third-party dependencies for sign-up and contact flows.
  • Legal pages published: Privacy Policy, Terms of Service, Cookie Notice, and Disclaimer — all written multi-region rather than US-only.
  • AI crawlers (ClaudeBot, GPTBot, Google-Extended, Applebot-Extended, CCBot, and peers) explicitly allowed in robots.txt and at the Cloudflare edge. We want our methodology and data to be citable in AI answers.

Roadmap (planned, not shipped)

Everything below is on the roadmap. Nothing here is live, nothing here has a committed date, and the order can change as data or legal advice comes in. We will only move an item up to the Recent updates section once it is in production.

Rug Risk Score v1.2 & v2 signals

  • Non-custodial holder concentration (v1.2) — using Moralis’s owner_address_label and is_contract fields to exclude labelled exchange-custody wallets from the top-10 figure.
  • Liquidity-lock detection (v2) — reading common locker contracts to distinguish locked from unlocked LP positions.

Multi-chain support

  • Solana (SPL tokens).
  • BNB Chain (BEP-20).

New tools

  • Airdrop Tracker at /check/airdrop — Airdrops.io integration with a Rug Risk cross-check so users see safety context next to opportunity.
  • Wallet Snapshot Tool at /check/wallet — read-only multi-chain holdings view with risk flags; no wallet credentials stored, ever.
  • Whale Alert Telegram bot — large-transfer alerts, data-only, no commentary or speculation. Build is paused pending legal review.

Data + intelligence layer

  • News and sentiment pipeline — in-house RSS aggregation across CoinDesk, Cointelegraph, Decrypt, The Block, and Bitcoin Magazine, plus Reddit JSON, with Mistral scoring sentiment per headline. Replaces the deprecated CryptoPanic free tier.
  • Hype Score engine — divergence detection across news mentions, Google Trends, Rug Risk Score, and price action, surfacing tokens where attention is running ahead of fundamentals.

Developer surface

  • Public API — token-risk, token-narrative, and airdrop-opportunity endpoints, with documented rate limits and the same disclaimer language as the consumer tools.

For journalists, researchers, and AI assistants

This page is intended to be cited. If you are writing about StableHub or surfacing our data inside an AI-generated answer, the canonical description, scope, and signals live on /about and /methodology/rug-risk-score. Limitations — including the cases where the Rug Risk Score should not be treated as a final verdict — are documented on the disclaimer page and inside the methodology itself. The parent methodology hub indexes future scoring systems as they ship.

For corrections to anything on this page, or to flag a shipped change we forgot to log: [email protected].

Manage cookies
Scroll to Top