[ANN][TESTNET] Kairos – Bitcoin-style PoW with quantum-fallback addresses

Kairos
A peer-to-peer electronic cash system built to last

Public testnet is live · Testnet coins have no value · No sale, no premine, no airdrop

What it is
Kairos keeps what has proven itself in Bitcoin (SHA-256d proof-of-work, UTXOs, the chain with the most work wins) and changes what seventeen years of operation have exposed. All of this is active from block 0; none of it is a later soft fork.

A security budget that never reaches zero. Emission decays smoothly with no halving cliffs, then settles at a permanent 0.6 KRS per block.
A burned base fee. A predictable per-byte fee, adjusted every block (EIP-1559 style). Burning it offsets the tail emission when the chain is busy.
A quantum fallback in every address. Each address commits to a Schnorr key and a hash-based Lamport key. If secp256k1 falls, a soft fork disables elliptic-curve spends and coins move under the hash-based key instead of being stolen or frozen.
The UTXO set is committed in every header (MuHash3072), so nodes can start from a recent snapshot verified against proof-of-work.
Difficulty adjusts every block (ASERT), with 2-minute blocks.
Merge-mining with Bitcoin (Namecoin-style AuxPoW), so a small chain isn't at the mercy of rented SHA-256 hash power.
Malleability-free txids, domain-separated Merkle trees, block height in the header, chain id in every signature (replay protection), optional transaction expiry.

Status: please read

This is an unaudited public testnet. Testnet KRS have no value and never will.
There is no mainnet, no token sale, no presale and no airdrop. Anyone selling "Kairos" is scamming you.
If a mainnet ever happens, it will be a fair launch with no premine and no presale, and only after the steps in LAUNCH.md: independent audits, a second implementation and at least six months of public testnet.
The reference implementation is about 3,300 lines of Python (standard library plus optional libsecp256k1), with 55 tests. It's built to be read and attacked, not to scale.

Join in three commands (Python 3.9+, macOS or Linux)

Code:

git clone https://github.com/Nimsalcade/kairos.git && cd kairos
python3 -m venv .venv && .venv/bin/pip install coincurve
.venv/bin/python -m kairos --testnet node --mine

Your node finds the network by itself through the built-in seed nodes in Frankfurt, Singapore and Miami. Type help at the prompt.

What I'm asking for

Run a node and try to break it: consensus bugs, crashes, sync failures, denial of service.
Review the consensus code (params.py, tx.py, block.py, auxpow.py, chain.py) and the whitepaper.
Pool operators: help test merge-mining through createauxblock / submitauxblock.
Honest criticism of the design, especially the tail emission, which I know is contested.

Bugs go to GitHub Issues. Security problems go through the private report on the repository's Security tab, not this thread, please.

Links
Code: https://github.com/Nimsalcade/kairos
Whitepaper: https://github.com/Nimsalcade/kairos/blob/main/docs/kairos-whitepaper.pdf
Known limitations: SECURITY.md · Mainnet criteria: LAUNCH.md

Disclosure
The design, whitepaper and reference implementation were drafted with the help of Claude, an AI model made by Anthropic. I deployed and run the testnet. Kairos is independent and not affiliated with or endorsed by Anthropic. Treat every line as unreviewed until people other than me have reviewed it; that's why I'm posting.

I'm Nimsalcade on GitHub. This thread will be linked from the repository README to confirm.