Changelog
What's changed in VERROW's checks, scoring rules, and pages. Scoring-rule changes are always also reflected on the Methodology page.
v1.20 — RugCheck-style safety checklist on token reports
August 2026
Site
- Token reports now open with a Safety checklist: scam flag, verification, mint, blacklist/pause, LP lock, liquidity, holder concentration, deployer holdings, sell simulation, tax, and proxy — each pass/warn/fail/unknown with a jump link to evidence.
v1.19 — Scored Telegram feed + deployer holdings
August 2026
Alerts
- Telegram posts new Radar tokens to a single channel with the VERROW risk score (any level). Soft filter: ≥$500 liquidity, ≥40% coverage, no explorer scam flag. Call Board / $100k admin DMs still only track the stricter Low-risk clean bar.
Scoring
- New finding when the attributed deployer still holds supply among analyzed holders: +20 above 20%, +10 for 10–20%.
v1.18 — Compare cinematic — critical differences first
August 2026
Site
- The Compare page now opens with a cinematic strip of per-token risk cards, then a Critical differences list (sell simulation, LP lock, scam flag, mint/blacklist, bytecode twins, and more) before the full metric table.
- Comparison table adds sell simulation, human-readable LP lock labels, and rapid-deploy rows. Page scan budget raised for multi-token cold scans.
v1.17 — Telegram alert reliability on serverless
August 2026
Alerts
- Fixed a failure mode where, without Upstash Redis, each serverless cold start re-ran Telegram bootstrap and seeded the whole Radar window as “already posted,” so the bot never sent clean-bar alerts and Call Board stayed empty.
- Without Redis, alerts now use a recent-only ephemeral mode (newly verified clean-bar tokens) instead of permanent silence. With Redis, first-run bootstrap seeds only tokens that already clear the clean bar — others can still alert later when they qualify.
- Cron radar-refresh responses now include richer Telegram diagnostics (mode, durableStore, lastError, cleanBarCandidates) plus env flags for telegram/redis configuration.
v1.16 — Deployer profile pages
August 2026
Site
- New /deployer/[address] profiles an EOA or factory as a deployer: recent created tokens with market cap/liquidity, rapid-deploy timing, and bytecode twins among its creates.
- Token reports link the attributed deployer address (rap sheet + Deployer History) to the dedicated deployer page.
v1.15 — Call Board — public clean-bar call scoreboard
August 2026
Site
- New /calls page lists tokens that cleared the Telegram clean bar, with entry market cap (recorded at alert time), live MC, tracked peak MC, and multiples. Sorted by peak multiple when available.
- Clean-alert tracking now stores market cap and risk score at alert time and updates peak MC during milestone polls and Call Board refreshes. Legacy tracked rows without entry MC still appear (liquidity shown) but multiples stay blank until new alerts record entry MC.
v1.14 — Share cards, deployer rap sheet, risk × market-cap history
August 2026
Reports
- Token report Open Graph share cards now surface sell-simulation, LP lock, market cap, and top findings — and Twitter/X previews use the large image card. Reports also offer a Share card download next to Share report.
- A compact Deployer rap sheet sits near the top of each report: attribution path, sibling tokens with peak MC, rapid-deploy, and bytecode twins — with a jump link to full Deployer History.
- Risk score history now records scan-time market cap (or FDV) alongside each snapshot and overlays it on the chart when available. Legacy history is migrated; older points show score only until new scans fill MC in.
v1.13 — Honeypot sell-simulation (Uniswap V2 + V3)
August 2026
Risk engine
- Token reports now run a simulated buy-then-sell (0.01 ETH) when the primary pool is a WETH-quoted Uniswap V2 or V3 pair — via Router02 (V2) or SwapRouter02 + pool fee() (V3). A successful buy followed by a reverted sell is scored as a critical honeypot-style finding; extreme effective taxes (≥25%) are flagged as high risk.
- V4 pools and non-WETH quotes are reported as unavailable rather than guessed. Simulation failures never fail the whole scan.
v1.12 — Deeper deployer attribution + sibling token market caps
August 2026
Reports
- When a token is created by a factory/launcher contract, Deployer History now attributes rapid-deploy, bytecode-reuse, and sibling-token checks to the creation-transaction sender (the account that triggered the factory) — not the factory address itself. The factory is still shown separately on the report.
- Other tokens recently created by that attributed deployer are listed with links to their VERROW reports, plus current DexScreener market cap and liquidity. ATH market cap is shown when DexScreener exposes it (often unavailable on Robinhood Chain).
v1.11 — Risk score history on token reports
August 2026
Reports
- Each fresh scan now records a Risk Score snapshot server-side. Token reports show a history chart and recent snapshots — how the score has moved across VERROW scans over time, not only since your last visit on this device.
- Identical scores within a short window are coalesced so busy tokens don't flood the series; history is kept for up to 90 days when Redis is configured.
v1.10 — Private $100k milestone DMs for clean alerts
August 2026
Alerts
- After a clean-bar token is posted to the public Telegram channel, VERROW tracks it privately. Every 15 minutes it checks DexScreener market cap; the first time a tracked token crosses $100k, only the admin receives a DM — the public channel is never notified.
- Requires TELEGRAM_ADMIN_CHAT_ID (your numeric Telegram user id after /start on the bot). Upstash Redis is strongly recommended so the tracked list survives across serverless instances.
v1.9 — Telegram clean bar — fewer, higher-quality alerts
July 2026
Alerts
- Telegram no longer posts every liquid or Critical Radar token. Alerts now require a clean bar: Low risk, at least $3k liquidity, ≥70% analysis coverage, and no Blockscout scam flag.
- Each post explains why it was selected ("Passed VERROW clean bar"). Silence in the channel means nothing cleared the bar — Radar on the site still surfaces the full feed.
v1.8 — Pro Telegram alert formatting
July 2026
Alerts
- Radar Telegram posts now use a clearer risk-colored header, surface Blockscout scam flags at the top, and list up to two other critical findings when present.
- Each alert includes inline buttons for the VERROW report, DexScreener, and the block explorer. When a token logo is available, the alert is sent as a photo with caption (falling back to text if the image can't be fetched).
v1.7 — Telegram Radar alerts
July 2026
Alerts
- New Radar detections can now be posted automatically to a Telegram channel after each refresh: contract address, risk score/level, liquidity, age, launchpad, scam flag (when present), and links to the VERROW report and DexScreener.
- To avoid spam, alerts only fire for tokens that already have liquidity, or that are Critical risk (e.g. Blockscout scam flag). The first run after enabling seeds silently so the current feed isn't dumped into the channel at once.
v1.6 — Radar filtering by risk level and liquidity, plus sorting
July 2026
Radar
- Filter the feed by risk level (Critical/High/Medium/Low) or by whether a token has liquidity yet, in addition to the existing launchpad filter.
- Sort the feed by newest, highest/lowest risk, or highest liquidity.
v1.5 — Real tax/limit reads, selfdestruct detection, and a liquidity-lock correctness fix
July 2026
Risk engine
- Buy/sell/transfer tax and max transaction/wallet limits are now read directly from contract state, instead of only flagging that a getter function exists.
- Added detection for a contract's ability to selfdestruct.
- Fixed a bug in Uniswap V3/V4 liquidity-lock detection: it could report a position's historical mint amount as still "held" even after that liquidity had since been fully withdrawn, in rare cases misclassifying a locked pool as not locked. Custodian analysis now verifies current on-chain liquidity for every sampled position and widens its search if the initial sample turns out to be fully withdrawn.
Performance
- Reduced duplicate work on a token's first (not-yet-cached) scan, and extended the scan's time budget so a slow-but-successful scan is no longer at risk of being cut off.
v1.4 — Radar filtering, proactive refresh, and token logos
July 2026
Radar
- Filter the Radar feed by launchpad label. The feed also now shows more tokens, refreshed more often.
- The feed is proactively refreshed in the background on a schedule, so visitors load an already-warm feed instead of triggering a live sweep themselves.
- Token logos now appear in the feed, preferring a logo read directly from the launch platform's own contract before falling back to market-data sources.
v1.3 — Token report redesigned as a glanceable dashboard
July 2026
Report design
- Replaced the linear risk score display with a circular gauge and a Quick Stats summary row (liquidity, holders, LP lock, max tax, contract verification) that link straight to their section.
- Report sections and individual findings are now collapsible, auto-expanding only when they contain a critical or high-severity result — so the page is scannable at a glance instead of a wall of text.
- Clicking a section link in the navigation now opens that section directly, instead of just scrolling to a collapsed one.
- Added the token's logo, live price, and website/X/Telegram links (when available) to the report header.
v1.2 — Block explorer scam flag now surfaces as a critical finding
July 2026
Risk engine
- Blockscout's own scam reputation flag for an address was fetched but never used. It's now checked on every scan and, when present, immediately produces a dedicated critical finding (+75) and a warning banner at the top of the report — independent of every other check.
v1.1 — Trading-control detection wired into scoring
July 2026
Risk engine
- A privileged trading-enable/disable function was already detected from verified source, but was not being scored or shown. It now produces a dedicated finding (+15) and appears on the Ownership and Permissions section.
v1.0 — Launch
July 2026
Core risk engine
- Contract verification, proxy detection, and permission-pattern analysis from Blockscout.
- Ownership and privileged-function detection (mint, blacklist, pause, fee/limit changes, upgrade authority).
- Holder concentration analysis excluding the token contract, burn addresses, and confirmed liquidity pools.
- Liquidity and market data from DexScreener.
- Deployer history and rapid-deployment detection.
- Liquidity lock detection across Uniswap V2, V3, and V4 pool designs, including known locker contracts.
- Deployer bytecode-reuse (clone) detection: flags redeployed template contracts from the same deployer.
- Evidence-based Risk Score (0–100) and Analysis Coverage (%), with a full Methodology page.
Radar: newest tokens, auto-scored
- A Radar page showing the newest verified token contracts on Robinhood Chain, each scored with the same risk engine as a normal report.
- Discovery is launchpad-agnostic — it watches the chain's newest-verified-contracts feed rather than a single factory or launchpad.
- A best-effort "via <launchpad>" label when the contract name matches a known pattern (e.g. PonsLaunch); a label only, never a filter.
User tools
- A holder concentration chart and an embedded DexScreener price chart on every report.
- A Watchlist stored locally in your browser, no account required.
- A score-change indicator showing how a token's Risk Score changed since your last visit, on this device.
- A Compare page for viewing two to four tokens side by side.
- A documented public API.
Discoverability & trust
- Sitemap, robots.txt, structured data, and dynamic per-token social preview images for better discoverability and sharing.
- A full Methodology page documenting every scoring rule, data source, and known limitation.
Found a bug or an incorrect finding?
VERROW never claims perfection — if a report looks wrong, send us a message with the token address and what you expected to see.