privacy policy
Last updated
01information we collect
Account data (only if you sign up). Creating an account requires an email address, a username, and a password. Authentication is handled by our infrastructure provider Supabase; passwords are stored hashed and are never visible to us. If you follow accounts while signed in, the list of accounts you follow is stored against your user ID so it syncs across your devices. Signing up also takes an invite code: we keep the code, who invited you, how you heard of sniff if you tell us, your time zone, and when you accepted the Terms.
Early-access requests. If you submit the waitlist form, we collect what you enter there (email, optional X handle, and your answers) to manage invitations. If you ask for an invite, we collect your Telegram handle and, if you give it, your X handle.
Automatic technical data. Like nearly every website, our hosting infrastructure (Vercel) produces standard server logs — IP address, user-agent, requested URLs, and timestamps — used for security, debugging, and abuse prevention. We do not build behavioral profiles from these logs. To stop floods, our servers also count requests from each IP address, and keep some of those counts in our database for about three days, with no content. For each account, we note the days it was used and how many pages and data calls it made on each, with no pages or addresses.
On-device storage (localStorage). Your theme preference and, when signed out, your follow list live only in your browser’s localStorage on your device. When you open your personal track feed, the follow list is sent with that request so we can build the feed. The application does not store it beyond serving the response (unless you are signed in and it syncs to your account as described above) — but because it travels in the request URL, it appears in the standard server logs described above for their short retention window (section 05). When you sign in, your session is kept in your browser’s localStorage and in cookies the site sets to keep you signed in; the cookie that renews your session lasts up to 400 days unless you sign out. The site also sets cookies for your time zone and for how some pages are laid out. None of these is used for advertising.
Subscription & billing data. If you subscribe, your payment card details go directly to the payment processor — Apple (App Store purchases) or Whop or Stripe (web purchases) — and never touch our servers. What we keep is your subscription state: which plan you bought, its status and renewal date, which store it came from, and the processor’s identifiers for the customer and subscription, so your purchase unlocks the Service on every device and support questions are answerable. We also record paywall events (which paywall you saw and what you did on it) to run the Service’s pricing experiments. Some memberships and the Data API’s credits are paid in SOL: we keep the amount, the time, the wallet that paid and the transaction’s signature, all of which are public on the Solana blockchain, so the payment can be matched to your account.
Product analytics (PostHog). We use PostHog, hosted in the United States, to understand how the Service’s own features are used — screens viewed, features tapped, subscription funnel steps, and application errors, tied to your user ID when you are signed in. This is our own product measurement, not advertising: no data is shared with ad networks and no cross-site tracking profile is built. PostHog may also record a replay of how a page is used (what was on screen, where you clicked and scrolled), with what you type into fields hidden. PostHog does not run on the chat’s page or on the page where you add a model key.
Visit counting (first-party). The site also counts its own page views on its own servers — a random visitor id kept in your browser, the page path without its query, the referring site’s domain, your country as reported by our host and whether you are on a phone, tablet or desktop, tied to your user ID when you are signed in — and keeps no IP address or browser string with them, sending nothing when your browser signals Do Not Track or Global Privacy Control.
What we do NOT collect: no payment card numbers, no brokerage or exchange connections, no government IDs, no precise location, no contacts, no advertising identifiers. We use no advertising SDKs or tracking pixels, and we set no advertising cookies. (Embedded media players on episode pages are third-party content and may set their own cookies — see section 04.)
02public-figure & creator information
The Service indexes publicly available content and records: public podcasts, streams, videos, posts, and newsletters published by creators, and public government records such as SEC filings and congressional financial disclosures. From these we display names, handles, avatars, quotes, summaries of expressed market views, and derived figures. On the crypto side the Service also reads public posts on X, Reddit, TikTok, FOMO and pump.fun, and public blockchain records such as what a wallet holds and trades. It never says who wrote a FOMO or pump.fun post, but it shows the public handle of a FOMO or pump.fun trader whose public calls it measures, with that record.
We process this information for journalistic, research, and indexing purposes on the basis of legitimate interest in making already-public financial commentary searchable and comparable. If information about you appears on the Service and you believe it is inaccurate, misattributed, or should be removed, contact legal@paste.trade — we review all requests promptly.
03how we use information
- to provide and operate the Service (build feeds, sync follows across devices, remember preferences);
- to process subscriptions — unlock what you paid for on every device, handle renewals and cancellations, and answer billing questions;
- to run the Data API: answer each call, charge its credits, keep each key within its cap, and show you your history;
- to run the chat when you use it: send your messages to the model provider you chose, and keep your chats if you ask it to;
- to measure and improve our own product (which features are used, where errors happen, how the subscription funnel performs);
- to secure the Service (rate limiting, abuse and fraud prevention, debugging);
- to manage the waitlist and send invitations you requested;
- to comply with law and enforce our Terms.
We do not use your personal information for advertising, we do not sell it, and we do not “share” it for cross-context behavioral advertising as those terms are defined under the California Consumer Privacy Act.
04who we share information with
- Service providers (processors): Supabase (database, file storage and authentication, which also sends sign-in emails), Vercel (hosting, CDN, server logs), PostHog (product analytics and page replays, US cloud), Whop and Stripe (web subscription payments), Apple and RevenueCat (App Store subscription payments and receipt validation), an email service that sends sniff’s own letters (such as the link that confirms an account deletion), and Google Fonts (font files are loaded from Google’s servers, which receive your IP address as part of serving them). These providers process data only to provide their services to us.
- AI assistants you connect: if you connect the sniff plugin, the assistant you use (ChatGPT by OpenAI, Claude by Anthropic, or another) receives the answers to the requests it makes on your behalf, and nothing about you. When it shows sniff’s coin card, your device loads the coin pictures from DexScreener and Mobula, which receive your IP address. Section 08 says what that covers. If you connect the Data API, your assistant also receives, with each answer, what the call cost and your credit balance after it. Section 09 says what that covers.
- Data suppliers: when you or your assistant call a Data API tool that reads live data, what the tool was asked (such as a token or wallet address, a search’s words, a handle or a web address) is sent to the data supplier that answers it. Your name, email address and account are never sent.
- Model providers you choose: if you use the chat with your own model key, your messages, the chat so far and what the chat’s tools read are sent to that key’s provider (Anthropic, OpenAI, xAI, Google or OpenRouter), which handles them under its own terms with you.
- Public blockchains and Telegram: a payment you make in SOL is recorded on the public Solana blockchain, where anyone can see the wallet that paid, the amount and the time. If you message us on Telegram, Telegram handles that conversation under its own privacy policy.
- Legal: we may disclose information if required by law, subpoena, or legal process, or to protect the rights, safety, or property of the Service, our users, or the public.
- Embedded content: episode pages embed players from the platforms the content lives on — YouTube (Google), X/Twitter, Instagram (Meta), TikTok, and Twitch (Amazon). When such an embed loads, that platform receives your IP address and user-agent and may set its own cookies under its own privacy policy, as it would if you visited the platform directly.
- Business transfers: if we are involved in a merger, acquisition, or asset sale, user information may transfer as part of that transaction; this Policy would continue to apply until changed.
Beyond the providers, suppliers, embedded platforms, picture hosts and services above, no other third parties receive your personal information.
05retention
- Account data: kept until you delete your account. Deleting your account (from your settings, confirmed by a link we email you) removes your profile and follow records, your Data API credits, keys, connections and call records, your chat keys and saved chats, and deletes your authentication user; residual copies in encrypted backups roll off on the backup schedule.
- Subscription records: kept while your subscription can still renew and afterwards as required for accounting and tax law; deleted with your account otherwise. Records of the SOL payments we received (the amount, the time, the wallet that paid and the transaction) are kept for our accounts after your account is deleted, without the link to your account.
- Data API and chat: your balance, keys, connections and the record of each call are kept until you delete your account. A connection that goes unused for 30 days ends. Your saved chats are kept until you delete them or your account, and a model key until you remove it or delete your account.
- Analytics events: retained in PostHog on its standard retention and deletable on request with everything else (contact below, or delete your account).
- Server logs: kept by our hosting provider, Vercel, for one day, then deleted. The per-address request counts in our database are deleted after about three days.
- localStorage: stays on your device until you clear it (browser settings) — we cannot delete it remotely.
- Waitlist entries and invite requests: kept until invitations conclude or you ask us to remove yours.
06security
We use industry-standard measures: TLS encryption in transit, hashed passwords, row-level security on our database, and least-privilege service credentials. Data API keys and connection tokens are stored only as fingerprints (hashes), and chat model keys and saved chats are stored encrypted. No system is perfectly secure, and we cannot guarantee absolute security — please use a unique password. If we learn of a breach affecting your personal information we will notify you as required by applicable law.
07your rights & choices
Everyone: you can use the Service without an account; you can clear on-device data in your browser; you can delete your account yourself from your settings at any time; and you can contact legal@paste.trade to access, correct, export, or delete data we hold about you. We respond to all requests within the time required by applicable law and never discriminate against you for exercising rights.
European Economic Area / United Kingdom (GDPR): where GDPR applies, our legal bases are contract (providing the Service you request), legitimate interests (security, indexing public information), and consent where required. You have rights of access, rectification, erasure, restriction, portability, and objection, and the right to lodge a complaint with your supervisory authority. We do not use automated decision-making that produces legal or similarly significant effects about you.
California (CCPA/CPRA): you have the rights to know, delete, correct, and port your personal information, and to opt out of sale or sharing — we do not sell or share personal information, so there is nothing to opt out of. We honor these rights without discrimination. The categories we collect are described in sections 01, 08 and 09; we disclose them only to the recipients in section 04.
Do Not Track / Global Privacy Control: we run no cross-site tracking at all. Our own visit counting sends nothing when your browser sends either signal; the rest of the Service, PostHog included, behaves the same whether or not these signals are present.
08the sniff plugin for ChatGPT and Claude
What it is. The sniff plugin lets an AI assistant you use (ChatGPT, Claude, or another that speaks the Model Context Protocol) ask sniff.trade for what it already holds: the coins it covers, a coin’s stored facts and analysis, attention by theme, chain reports, the bot’s track record, and published research. It is read-only. It fetches no live prices, starts no analyses, places no trades, and carries no advertising.
No sign-in. Every tool is open to everyone. The plugin asks for no account, no password and no email address, sets no cookies, and does not learn who you are.
What a request carries. Your assistant decides when to use a tool and writes what the tool is asked. The request reaches our server from your assistant’s provider (OpenAI or Anthropic), not from your device, and carries:
- the tool’s name and what it is asked: a ticker, a coin’s name or contract address to search for (up to 120 characters), a chain, a theme, a time window or period, how many results to return and from which one, or a research piece’s address;
- the details every web request carries: the sending computer’s IP address (usually your assistant provider’s, not yours), its user agent, and the time;
- what your assistant chooses to attach to each call. ChatGPT attaches your language, a rough location (city, region, country, time zone and approximate coordinates), a hint of the app you use, and anonymous ids for you, your conversation and your organization. sniff does not read, use or keep any of these.
A request never carries your conversation, your name, your email address or your account with the assistant. Put nothing personal in a question you want sniff to answer, since your assistant may pass words from it to a search.
What we keep, and for how long.
- Request logs. Our host, Vercel, logs each request: the time, the sender’s IP address and user agent, the path (/api/mcp), the result status, and one line naming the tool that was called. We never log what a tool was asked. Vercel keeps these logs for one day, then deletes them.
- Error notes. When a tool fails, the same log keeps a short note with an error code and the reason, never what the tool was asked. It is deleted after one day with the rest.
- Abuse limits. To stop one sender flooding the service, our server counts requests per sending IP address in its memory for one minute, and our host’s firewall does the same at its edge. The server also counts all plugin requests together, per hour, in our database, with no content, so that no one can run up our costs, and keeps each sending IP address’s share of that hour there too, deleted after about three days.
- Nothing else. What a tool is asked, and what your assistant attaches to the call, is held in memory only while the answer is built. We store no copy of it, no copy of the answer and no conversation, and we build no profile from plugin use.
What we use it for. To answer the request; to keep the service working and safe (debugging, rate limits, blocking abuse); and to count how often each tool is used, so we know what to improve. We do not use plugin requests for advertising, we do not sell them, and we do not use them to train AI models.
What the assistant receives. The answers to the requests it makes: coin facts and stored figures with their dates, theme and chain reports, the bot’s past results, research text, and the addresses of coin pictures. The answers hold nothing about you. They become part of your conversation with the assistant, which OpenAI, Anthropic or your assistant’s provider handles under its own privacy policy, not this one.
Who else receives data.
- Vercel hosts the plugin’s server and keeps the request logs above for one day.
- Supabase runs our database, which the tools read from and which holds the hourly count above.
- When your assistant shows sniff’s coin card, your device loads each coin’s picture straight from DexScreener’s image server (cdn.dexscreener.com) or Mobula’s (metadata.mobula.io). Those services receive your IP address and user agent as they serve the picture, under their own privacy policies. The card loads nothing else.
- Your assistant’s provider (OpenAI, Anthropic, or another), as above.
No one else receives plugin data, and we never sell it.
Your choices.
- Use it without an account: there is nothing to sign up for, and nothing stored about you to delete.
- Choose what you ask: the plugin only answers what your assistant sends it.
- Turn it off: remove or disconnect sniff in your assistant’s plugin, app or connector settings, and the assistant makes no further requests to sniff. Your conversation history stays with your assistant, under its own controls.
- Ask us: write to legal@paste.trade, or use the support page, with any question or request about plugin data. Request logs are deleted after one day and the per-address counts after about three days, so after that there is nothing left to find.
09the Data API, its connector and the chat
What it is. The Data API lets you, your own code, or an AI assistant you connect (Claude, ChatGPT, or another that speaks the Model Context Protocol) read crypto data from sniff: live token and wallet data, what traders wrote on FOMO and pump.fun, sniff’s themes, recaps and research, and public posts on X, Reddit and TikTok and web pages, read live. It only reads: it places no trades and moves no money. Each call spends credits you bought in advance.
Signing in. An assistant can list the tools without signing in. To call one, you sign in to sniff and press Allow, and that assistant gets a connection of its own, which it can use until you disconnect it or it goes unused for 30 days. Your own code uses an API key instead, which sniff shows you once.
What a call carries. The tool’s name and what it is asked (such as a token or wallet address, a search’s words, an X handle or a web address), your connection or key, and the details every web request carries: the sending computer’s IP address (for an assistant, usually its provider’s, not yours), its user agent, and the time. Your assistant decides what to send; a call never carries your conversation.
What we keep.
- Credits and calls. Your balance, and a record of each call: the tool, what it cost, the time, and which key or connection made it. We store no copy of what a tool was asked or what it answered; the server holds them in memory while it answers, and for ten minutes when your code asks it to answer a retry the same way.
- Keys and connections. For a key, its name, its last four characters and a fingerprint of it, never the key itself. For a connection, the assistant’s name and web address, a fingerprint of its tokens, what it has spent this month, and when it was made and last used.
- Purchases. What you bought, and the SOL payment that paid for it, as section 01 says.
- Logs and limits. Our host, Vercel, logs each request (the time, the IP address, the user agent, the path and the result) and one line naming the tool and its cost, never what it was asked, for one day. Counts per account, per key or connection and per IP address, which keep anyone from crowding out the rest, are kept in memory and in our database for about three days.
Who else receives data. A tool that reads live data sends what it was asked to the data supplier that answers it, never who you are. Your assistant’s provider receives each answer, with what the call cost and your balance after it, and handles it under its own privacy policy, not this one.
The chat. If you use the chat in sniff with your own model key, the key is kept encrypted on our servers and opened only to send a step of your chat to that key’s provider, which receives your messages, the chat so far and what the chat’s tools read. While Keep chats is on, your chats are saved encrypted with your account; while it is off, nothing more is saved. We keep counts of what the chat used, to hold each account to its limits.
What we use it for. To answer each call and charge for it, to keep your caps, to show you your history, to keep the service working and safe, and to count how often each tool is used. We do not use Data API calls or chats for advertising, we do not sell them, and we do not use them to train AI models.
Your choices. Disconnect an assistant in your settings under Connected assistants. Revoke a key on the Data API page of your account. Remove a model key where you added it, and delete a saved chat in the chat. Or delete your account, as section 05 says.
10international transfers
The Service’s servers and database are hosted in Ireland, in the European Union, and our providers may process data in other countries, including the United States, where PostHog is hosted. Where required, transfers rely on appropriate safeguards such as standard contractual clauses offered by our processors. By using the Service you understand your information may be processed outside your own country, including in the United States.
11children
The Service is not directed to children and, per our Terms, users must be 18 or older. We do not knowingly collect personal information from anyone under 13 (or the equivalent minimum age in your jurisdiction); if you believe a child has provided us information, contact legal@paste.trade and we will delete it.
12changes & contact
We may update this Policy as the Service evolves. Material changes will be reflected by updating the date at the top; significant changes will be highlighted on the Service. Continued use after a change means you accept the updated Policy.
Contact for anything in this Policy — including access, correction, deletion, creator/public-figure removal requests, and complaints: legal@paste.trade