El Bukis: What It Was And Why It Disappeared
El Bukis was a Chrome extension built around 2018 that let you browse blockchain data through a privacy-focused lens while simultaneously running as a lightweight node for the Ethereum ecosystem. The creator, operating under the handle "buki," shipped it with zero marketing budget and no public roadmap. It gained traction purely through Reddit threads and Telegram groups where early adopters discovered that it could sign transactions directly from the browser without touching a third-party RPC endpoint. The thing most people got wrong about El Bukis is that it wasn't really a blockchain explorer. It was a client-side signing harness that happened to render transaction histories in a way that looked like a block explorer. You'd paste an address and it would fetch the data through your own local Geth instance or a public endpoint you configured yourself. The distinction mattered more than it should have, and it's probably the reason it died.
The Rise And Fall Of El Bukis The Mysterious Chromebased Blockchain Browser That Won The Internet
It won the internet in the same way most niche crypto projects do — someone with a decent following tweeted about it, GitHub stars climbed past three thousand in two weeks, and then every copycat project tried to build the same thing with a worse product. The original repo was archived in late 2019. No statement was posted. The domain went parking. The Chrome Web Store listing was quietly removed by the owner. I ran El Bukis for about eight months, roughly from March to November of 2019. Here's what it actually looked like to use day to day. You installed the extension, pointed it at your own Ethereum node or an Infura key you generated specifically for it, and then you could sign transactions without ever leaving the browser. No MetaMask. No hardware wallet dongle plugged into a desktop you kept for that purpose. The extension held a local encrypted key store encrypted with a passphrase you set on first launch. Transactions were constructed and signed locally before broadcast. That was the selling point.
The UI was ugly. It was built on a framework I never managed to identify — it didn't look like React, Vue, or Angular. Probably something vanilla with a custom template engine. The design treated every screen like a terminal output. You'd see raw hex strings, gas prices in gwei displayed with zero decimal padding, and transaction hashes that you could click to open in Etherscan if you had one configured. For people who were already running their own nodes, this was exactly what they wanted. For everyone else, it was a friction problem that the project never solved. There was no built-in fallback to public RPCs that worked reliably. If you didn't have your own node or a valid API key, El Bukis showed you a blank page and a status indicator that just said "disconnected" in gray text. No error message. No guidance. Just gray.
Get the Full Details
How It Actually Worked Under The Hood
The architecture was straightforward enough that someone competent could have forked and improved it within a weekend. The extension injected a JavaScript object into the page context called BukisProvider. Any dApp that supported EIP-1193 could talk to it directly. It acted as a JSON-RPC bridge between the browser and your configured Ethereum endpoint while intercepting wallet-related calls to handle signing locally. The signing flow went like this: you initiated a transaction through the extension popup, reviewed the data, entered your passphrase, and the extension signed it using the local keystore. The signed transaction was then broadcast through whichever endpoint you had configured. Everything happened client-side. No telemetry. No analytics. The creator was serious about the privacy angle. That seriousness became a liability. Because there was no analytics and no user feedback loop, bugs accumulated silently. I filed two issues on the GitHub repo in August 2019. One was about gas estimation returning zero on contract interactions with complex logic. The other was about the key export feature producing malformed files when the passphrase contained certain Unicode characters. Neither was addressed before the archive date.
Here's the specific problem I ran into that made me drop it. I was trying to interact with a DeFi protocol that was doing multi-signature logic through a proxy contract. El Bukis would construct the transaction fine, but when it tried to estimate gas, it would return a value that was roughly sixty percent of what the actual execution required. The transaction would go through the signing flow without complaint, hit the network, and revert with an out-of-gas error. I spent about three days digging through the source to find where gas estimation happened. The estimation logic was calling eth_estimateGas directly on your configured endpoint. For contracts with internal state mutations or call chains across multiple addresses, certain nodes would return optimistic estimates that were too low. The workaround I used was simple but annoying: I'd estimate gas through the extension, then multiply the result by 1.6 and hardcode that value into the transaction. It wasn't elegant. It worked consistently enough that I stuck with it for the remaining months I used the tool.
Why It Failed
There are a few overlapping reasons. The primary one is that El Bukis arrived at the exact moment MetaMask was transitioning from a niche tool to the default wallet for the entire Ethereum ecosystem. MetaMask offered the same client-side signing capability with a significantly better onboarding experience, built-in fiat on-ramps, and a growing ecosystem of dApp compatibility. El Bukis offered nothing MetaMask didn't except the philosophical commitment to running your own node. Secondary reason: the creator clearly valued privacy and minimalism over sustainability. There was no donation address visible in the repo. No funding round. No business model. The project existed because one person wanted it to exist. When that person moved on, the project moved with them. Tertiary reason is something I noticed but never saw discussed in any community thread. El Bukis did not support Ethereum Classic or any of the other EIP-155 chain variants that had meaningful ecosystems at the time. It was Ethereum-only. For a project built around self-custody and node-running, that's a strange limitation. The codebase had chain-id handling in it, but it was hardcoded to 1. Adding support would have been a matter of weeks of work. No one did it.
What Survived
The GitHub repository is still archived and readable. If you want to study how the provider injection worked or how the local encryption layer was implemented, the code is there. Several smaller projects have borrowed pieces of the architecture. The way El Bukis handled passphrase derivation using scrypt with a fixed salt pattern ended up being referenced in at least two follow-on tools that appeared in early 2020. There is no download link that works today. The Chrome Web Store listing has been removed. If you find a cracked version on some third-party site, don't use it. The extension interacted with private keys. A modified version could harvest anything it touched.
What To Use Instead Today
If you want client-side signing with a self-hosted node, RainbowKit paired with a local Geth or Erigon instance gets you nine tenths of what El Bukis offered, with active maintenance and a much larger support community. If the node-running part isn't actually important to you and you just want a clean signing experience without telemetry, Rabby Wallet has been quietly accumulating features that make it the closest spiritual successor in terms of philosophy, even if the product is fundamentally different. El Bukis worked for a specific slice of users at a specific moment. It was never going to scale beyond that slice. The people who used it successfully knew exactly why they were using it and accepted the tradeoffs. The rest of us watched from the side and learned from what it did right before it disappeared.
