why i killed my own crypto project (and built something that could actually live)

I built Health Protocol thinking if you gave people privacy, incentives, and proof, everything would just work. I shipped the infra and wrote the math with Claude’s help. I modeled every outcome using Monte Carlo sims. Every time, the same ending. Whales won, bots drained the pool, real users quit. The DAO was a graveyard. Nothing felt new. It was the same fate as every corporate wellness app, insurance “incentive,” and NFT move-to-earn. STEPN made you buy $1,200 sneakers for a shot at a token jackpot. By the end, bots were farming, prices tanked, and most users left holding nothing but another Discord full of bad memes.

I did the research. I looked at everything: DeFi, behavioral econ, read Richard Dawkins and thought about how casinos and governments try to buy good behavior. None of them solved the core problem: how do you reward people for what actually matters and keep privacy as a non-negotiable? The old Health Protocol yellowpaper documented exactly how it would fail.

arishem, eternals, and building for life, not stasis

The moment I gave up on the old protocol was sometime after midnight, Marvel’s Eternals on in the background, Arishem explaining why universes are born to die and recycle. It clicked. The problem wasn’t just incentives, it was immortality: THE VAMPIRE PARADOX! If you build for stasis, the system probably gets gamed and dies a boring death. If you build for life with decay, adaptation, feedback, etc, then it turns out you get a protocol that might actually live forever. Arishem: cosmic midwife, blockchain muse.

value that vanishes, and the birth of the reservoir

Every protocol I looked at was built on the same dead logic: make a token, freeze value, pray nobody dumps. That doesn’t work. Real biological systems move. In Autophage, tokens decay. You can’t save indefinitely though Wellness Vaults (think HSAs but on-chain) are a thing. You either have to use tokens you earn right away or set them aside for specific goals like how college funds and HSAs work, and if you ghost, your balance fades out.

The first person who called this out as bullshit was my mom. She said, “money isn’t supposed to disappear.” I tried to explain decay, game theory, and why hoarding ruins incentives. She said, “let it all flow back to everyone else, like a reservoir.” She was right so that’s exactly what I built; the missing piece. The Reservoir collects all decayed tokens and recycles them as system fuel. It pays future rewards, covers healthcare payouts, and backs the protocol’s safety net. My mom named it, shout-out to you Momma!

So you earn tokens, those tokens decay at different rates depending on how you earned them, i.e., through what exercise or activity. You can spend them lots of ways, but if you don’t they simply return to The Reservoir, funding other people’s health emergencies or yours. The Reservoir remembers 100% of the value you contribute through token decay and works like how “social security” is supposed to, except it doesn’t run out and is USDC backed through platform and marketplace fees.

game the system, then evolve

You earn tokens as mentioned before through health-related activities that include lots of things like sleeping, running, ab workouts, therapy session, STI checks, etc, really any health activity you can design a verification proof off of. Those activities are private to you and they should be, so the system implements a proof generation system. This gets complicated, but the gist is you access an App on the network like a theoretical status.health for STI test history verification, that App then generates a proof you did a thing, and sends it to the network. The network then pays you for your proof and stores it under your ProfileID which is cryptographically separated from your identity.

Proofs that are generated are not just trophies, you can actually turn them into products you sell, and even cooler, the protocol expects you to game it. Thanks to my buddy Dakota, who explained why Path of Exile works: you can always evolve your build and break it in new ways, and it stays fair. That’s how genetic adaptation landed in Autophage. Burn tokens, unlock new traits, specialize. If you’re a marathon runner, show your trait stack and earn multiples on running activities. If you care about privacy, share less, and still get paid. The more you adapt, the better your rewards. Apps and businesses pay to issue, verify, and use the protocol. Users just access and build reputation. All fees are in USDC. No protocol token games. If you spam, you get cut out. If you contribute, you get paid.

The protocol uses zkVM orchestration for proof verification - basically allowing apps to verify you did something health-related without ever seeing your actual data.

onlyfans for proof

The protocol flips health data ownership just like OnlyFans did for the porn industry. Not only can apps help you verify activity, but they can also help you sell health proofs you generate through the marketplace. Buyers could include individuals or businesses for things like medical research or post-hookup should a partner get anxious about your testing history. Anonymous marathon proof sells for $10-20 baseline. Reveal your ProfileID so buyers can follow your journey, earn 50-100% more. Show customizable traits your spec’d in proving you’re an endurance specialist, command $50+ per proof. Therapy consistency, workout streaks, STI tests, sleep quality, whatever you track becomes a potential product you can benefit from.

The marketplace has three layers. Individuals buy proofs from each other to study real health journeys or verify fitness, kind of like peacocks and their feathers. Someone training for their first marathon buys experienced runners’ proofs to understand pacing. Dating apps verify STI status without storing medical records. Research organizations purchase anonymous population data at scale. A university studying exercise patterns might buy 10,000 anonymous running proofs for $50,000. Every transaction protects privacy through zero-knowledge separation. Buyers get verification without identity. Sellers get paid without exposure.

You control the terms. List anonymous and stay completely private. Build a following around your ProfileID without revealing your name. Display specialized traits when expertise pays. The protocol takes 12% to fund The Reservoir. Apps that verified your activity get 5%. You keep 88% in USDC. Cash for health receipts. Privacy settings determine price. The market decides value. You own what you earn.

proposals and upgrades as a real job

The protocol treats development like a real market. You can ship any improvement that makes the system better. New proof mechanisms for different health activities, stronger privacy circuits, reputation algorithms that catch gaming, incentive curves that balance rewards. When apps adopt your code and it processes real health data, you capture value through integration fees, performance bounties, usage royalties, and security rewards. Early simulations show active developers earning $30-90k annually just from base activity, but that’s conservative math. The real opportunity is building critical infrastructure that thousands of apps depend on, like npm packages that print money. No voting, no committees, no “community calls” where nothing happens. Pure market dynamics: your code gets used, you get paid. It doesn’t, you don’t. The protocol needs everything from better zkVM orchestration to ML models for health insights to integration bridges for existing systems. Each improvement creates compound value and compound earnings. It’s an economic system where protocol improvements translate directly to developer income indefinitely as the network scales.

Anyone can propose improvements by staking 100-5,000 tokens on specific health metrics. The protocol runs controlled experiments for 30-180 days. Hit your 5% improvement target and get your stake back plus USDC bonuses proportional to impact. Miss and The Reservoir takes your tokens. Good improvers build reputation, stake more, run bigger experiments. Stack wins and you’ve got a real career improving public health infrastructure while getting paid in real money. The network adapts through these continuous experiments, evolving based on what actually works.

the usdc loop

Apps pay $0.10-0.20 monthly per user for verification rails. Enterprises drop $20-100 per employee wellness check. Users sell health proofs to each other with the marketplace taking 12%. All USDC flows to The Reservoir. The split: 10% for protocol development, 90% for healthcare settlements and reserves. Users earn USDC selling anonymous marathon proofs for 10 bucks or specialized genetic traits for 30. Apps compete on features while paying base infrastructure fees. Healthcare providers submit tokens and get settled in USDC at metabolic prices (essentially dynamic pricing based on the actual metabolic cost of healthcare activities). Developer improvements that get adopted earn integration fees and usage royalties. More network activity creates more fees, deeper reserves, better healthcare coverage. Your lifetime token contributions track your healthcare access rights. Decay funds the community, fees fund the infrastructure, proofs fund the users. Everything cycles through The Reservoir keeping value moving, never pooling.

receipts, math, and where to break things

The yellowpaper contains all the real math - every incentive curve, privacy proof, and system invariant. There’s also a patent filed covering the core mechanics. The protocol documentation includes code snippets, use cases, and simulation frameworks. If you can break it, send through a report to info@autophage.io.

sims don’t lie

The Health Protocol died every time in Monte Carlo simulations. Whales accumulated or bots farmed rewards or real users quit. Classic crypto death spiral. I killed it and built Autophage with different physics. Tokens that decay prevent hoarding. Genetic traits reward specialization over capital. The Reservoir recycles value instead of letting it pool. Ran thousands of simulations with every attack vector I could imagine. Bot armies, whale cartels (lol), bank runs (omg), you name it (I prob tried it). The protocol bent but didn’t break. Value stayed distributed because decay forces circulation. Users specialized into sustainable niches. The marketplace found equilibrium prices. The Reservoir maintained reserves through fee flows. Not perfect, but the numbers finally worked. Sometimes you have to let something die to build something that can live.

the last thing i’ll say about rocks

Crypto built itself on hodling. Diamond hands. Never sell. Rocks don’t need food, water, or healthcare. We do. The protocol’s meme captures it: “hodl? lol. not here.” Tokens decay because life does. Value moves or dies. The protocol name Autophage comes from autophagy - the biological process where cells eat their own damaged parts to survive. For those hunting deeper secrets, Charlotte holds the key.

What if our deepest economic assumption that value should persist forever is perfectly inverted?