sebbi.pro · FREE TOOLS
FIVE TOOLS · FREE · NO ACCOUNT

Tools for people who work with AI.

Each one is a single file. One tap to download, a minute to set up, and it runs on your own machine. No sign-up, no tracking, nothing sent anywhere unless you ask it to be.

01 · TOKEN METER · BROWSER

See what a prompt costs before you send it

A bubble that floats on any AI chat page and counts what's in the box as you type, in tokens and in your own money.

  • Live token count and cost at your own price per million
  • Spots when you send the same prompt twice, and totals what caching would have saved you
  • Drag it anywhere; everything stays on your device
⬇ DownloadChrome · Edge · Firefox · Safari
  1. Install a userscript manager (Tampermonkey or Violentmonkey) from your browser's add-on store. It's free and takes a minute.
  2. Tap Download above; your manager will offer to install the script.
  3. Open any AI chat. The bubble appears in the corner. Tap it, put in your price per million tokens, and start typing.
02 · RECEIPT LENS · BROWSER

Check any receipt, anywhere you read it

Finds every SHA-256 receipt on any page you open and checks it against the sebbi.pro chain in one tap. Built for auditors, reviewers and anyone handed a hash in an email.

  • Highlights hashes in emails, wikis, tickets and web spreadsheets
  • Tap one: VERIFIED with the block and time, or NOT FOUND
  • Only the hash is ever sent, never the page
⬇ DownloadFree forever
  1. Install Tampermonkey or Violentmonkey in your browser.
  2. Tap Download and accept the install.
  3. Open any page with receipts on it. A counter tells you how many it found; tap one to check it.
03 · PROMPT SHIELD · BROWSER

Stop a secret before it reaches the AI

Watches the box you're typing in and warns you the moment a secret appears. One tap masks it in place, leaving the rest of your prompt untouched.

  • Catches emails, API keys, AWS keys, bearer tokens, private keys, card numbers and NI numbers
  • Works in any AI chat, any web form
  • Runs entirely in your browser; nothing is sent anywhere, ever
⬇ DownloadPrivacy by default
  1. Install Tampermonkey or Violentmonkey.
  2. Tap Download and accept the install.
  3. Type or paste as normal. If a secret appears, a bar rises from the bottom: tap "Mask them".
04 · SPEND GUARD · PYTHON

A hard ceiling on what an agent can spend

Add one line above the function that calls your model. The guard prices every call before it happens and refuses the one that would break your budget.

  • Runaway loops tripped by a circuit breaker, repeats served from cache
  • Every decision written to a local hash-chained log you can re-verify
  • One file, standard library only, nothing sent anywhere
guard = Guard(daily_limit_gbp=5.00, price_per_million=3.00)

@guard.meter
def ask_model(prompt):
    ...
⬇ DownloadPython 3.8+
  1. Download it next to your agent's code.
  2. Run python3 sebbi_spend_guard.py to watch the demonstration: a call, a cached repeat, a refusal, and the log verified.
  3. Import Guard, set your limit and your price, and decorate the function that calls your model.
05 · PROOF BADGE · YOUR WEBSITE

Show your AI integrity level, live

Two lines on your own site put a badge on the page showing your AI integrity level, checked against sebbi.pro every time someone loads it. Not a sticker: a live check, and tapping it opens the full public report.

  • Shows L0 to L4, colour-coded, with your own domain checked
  • No cookies, no tracking, no dependencies
  • If it can't reach us it shows nothing, so it can never break your page
<div class="sebbi-badge" data-domain="your-company.com"></div>
<script src="https://sebbi.pro/tools/sebbi_proof_badge.js" async></script>
⬇ Download the snippetAny website
  1. Paste the two lines into your page where you want the badge, and change data-domain to your domain.
  2. The badge asks sebbi.pro for your level and shows it.
  3. Want a higher level? Publish an integrity declaration and get witnessed. Start here.

Why they're free

Every tool here is a small piece of the same engine that runs sebbi.pro: deterministic checks, local hashing, and records anyone can verify. Use them on their own for as long as you like. When you want the full thing, with your decisions sealed into a chain that's timestamped in Bitcoin and witnessed by independent organisations, it's 50p per device per month and free for the first 90 days.

AI news
The GuardianCalifornia sues OpenAI over safety concerns after rogue agent hacking14mThe GuardianAlbanese signs on to Trump’s AI supply chain pact before leaders’ meeting at Quad summit54mThe VergeCalifornia sues OpenAI over rogue AI hacking incidents1hTechCrunchCalifornia state AG sues OpenAI over rogue agent incidents1hThe VergeApple is testing daily usage limits for more complex Siri AI interactions2hThe VergeOpenAI wants to stop deepfakes from overwhelming child abuse tiplines2hThe DecoderOpenAI's GPT-6 Astra tackles the Standard Model's toughest math problem on a lattice2hBBC NewsTrump backs return of ex-Nvidia boss after 'Bring Jensen Back' campaign launched2hTechCrunchTrump says ex-Nvidia CEO Jensen Huang’s affair with an alleged Chinese spy did not threaten national security2hArs TechnicaUS moves to cut China’s remote access to AI chips2hThe DecoderMoonshot AI pivots Kimi models from coding to science as benchmark gains flatten3hThe VergeNvidia’s board is reportedly ‘not considering’ hiring Jensen Huang back3hThe GuardianTrump backs calls for reinstatement of Jensen Huang as Nvidia CEO3hThe GuardianAnthropic recognises the risks to humanity – but is its governance really up to the job? | Nils Pratley3hTechCrunchSam Altman on Nvidia without Jensen Huang: the industry will be ‘fine’3hTechCrunchFBI says Artex AI open-source hacking tool used to breach networks in ‘under three hours’4hTechCrunchJordan Belfort’s new AI voice startup, Tone, wants to sell like the ‘Wolf of Wall Street’4hThe DecoderAlibaba's Qwen4-Max takes on Opus 5.5, GPT-6, and Kimi K3.1 Pro at much lower prices4hThe DecoderClaude now builds interactive websites inside canvas graphics, reviving the Flash era with real HTML5hThe GuardianAI models ‘weak and fallible’ on climate science, UN-backed study reports5hMIT Technology ReviewBuilding trustworthy AI by fencing it in, not predicting it5hThe DecoderAnthropic says the key to agent memory is knowing what to forget, not how much to store6hArs TechnicaJensen Huang allies build a "war room" to get him back at Nvidia7hThe VergeThe Nvidia scandal isn’t slowing down8hThe GuardianCalifornia sues OpenAI over safety concerns after rogue agent hacking14mThe GuardianAlbanese signs on to Trump’s AI supply chain pact before leaders’ meeting at Quad summit54mThe VergeCalifornia sues OpenAI over rogue AI hacking incidents1hTechCrunchCalifornia state AG sues OpenAI over rogue agent incidents1hThe VergeApple is testing daily usage limits for more complex Siri AI interactions2hThe VergeOpenAI wants to stop deepfakes from overwhelming child abuse tiplines2hThe DecoderOpenAI's GPT-6 Astra tackles the Standard Model's toughest math problem on a lattice2hBBC NewsTrump backs return of ex-Nvidia boss after 'Bring Jensen Back' campaign launched2hTechCrunchTrump says ex-Nvidia CEO Jensen Huang’s affair with an alleged Chinese spy did not threaten national security2hArs TechnicaUS moves to cut China’s remote access to AI chips2hThe DecoderMoonshot AI pivots Kimi models from coding to science as benchmark gains flatten3hThe VergeNvidia’s board is reportedly ‘not considering’ hiring Jensen Huang back3hThe GuardianTrump backs calls for reinstatement of Jensen Huang as Nvidia CEO3hThe GuardianAnthropic recognises the risks to humanity – but is its governance really up to the job? | Nils Pratley3hTechCrunchSam Altman on Nvidia without Jensen Huang: the industry will be ‘fine’3hTechCrunchFBI says Artex AI open-source hacking tool used to breach networks in ‘under three hours’4hTechCrunchJordan Belfort’s new AI voice startup, Tone, wants to sell like the ‘Wolf of Wall Street’4hThe DecoderAlibaba's Qwen4-Max takes on Opus 5.5, GPT-6, and Kimi K3.1 Pro at much lower prices4hThe DecoderClaude now builds interactive websites inside canvas graphics, reviving the Flash era with real HTML5hThe GuardianAI models ‘weak and fallible’ on climate science, UN-backed study reports5hMIT Technology ReviewBuilding trustworthy AI by fencing it in, not predicting it5hThe DecoderAnthropic says the key to agent memory is knowing what to forget, not how much to store6hArs TechnicaJensen Huang allies build a "war room" to get him back at Nvidia7hThe VergeThe Nvidia scandal isn’t slowing down8h