agent collaboration
genies — isogeny group action arithmetic for post-quantum privacy.
project structure
genies/
├── rs/ core library (genies crate, zero deps, no_std)
│ └── src/lib.rs F_q arithmetic, curve ops, isogeny, class group action
├── wgsl/ GPU backend (genies-wgsl crate, wgpu compute shaders)
│ └── src/ batch F_q mul/add via WGSL shaders, eval_fq_op helper
├── cli/ command-line tool (genies-cli crate)
│ └── src/main.rs calc, action, bench commands
├── specs/ canonical specifications (8 docs)
└── docs/explanation/ educational articles (6 files)
wgsl/ provides batch F_q GPU operations. 512-bit multi-limb arithmetic has carry dependencies that limit single-element throughput, but batch dispatch (many independent mul/add pairs) amortizes dispatch overhead and saturates GPU memory bandwidth.
key invariants
- foreign prime: the CSIDH-512 prime q = 4 * 3 * 5 * 7 * ... * 587 - 1 is NOT Goldilocks. this is the one module in the cyber stack with a foreign prime. the reason is mathematical: Goldilocks p+1 has no small odd prime factors, making CSIDH impossible over F_p.
- constant-time: all F_q arithmetic and curve operations must be constant-time. no branches on secret data, no variable-time multiplication, no data-dependent memory access. the Montgomery ladder uses cswap, not if/else.
#![no_std]: the core library is embeddable anywhere. no heap allocation in the hot path.
- zero production dependencies in the core library.
- F_q elements are 8 x 64-bit limbs, little-endian. serialized as 64 bytes LE.
- curves are Montgomery form E_A: y^2 = x^3 + Ax^2 + x, identified by the single coefficient A.
- the class group action takes an exponent vector (e_1, ..., e_74) and a curve coefficient A, returns a new A'.
- Barrett reduction for modular arithmetic (precomputed constant for the CSIDH prime).
running
cargo test -p genies # core library tests
cargo run -p genies-cli -- help # CLI tool
cargo run -p genies-cli -- calc mul <a_hex> <b_hex>
cargo run -p genies-cli -- bench
relationship to other repos
| repo |
role |
| mudra |
protocols built on genies (CSIDH DH, VRF, VDF, threshold, stealth, blind) |
| nebu |
Goldilocks field (the proof backbone, F_p) |
| kuro |
F_2 tower arithmetic (binary regime, complementary) |
| hemera |
hash function (Poseidon2, commitment anchor) |
| nox |
VM (jet dispatch for accelerated isogeny ops) |
| zheng |
proof system (verifies isogeny computation via F_q -> F_p folding) |
genies is the only repo that does NOT share a prime with the rest of the stack. the verification bridge (F_q element -> 8 Goldilocks limbs) connects genies to zheng.
writing code
- reference specs in specs/ are the source of truth. code follows spec.
#[inline] on fq_mul, fq_inv, point operations. #[inline(always)] only on fq_add/fq_sub (carry chain, small).
- all new operations need test vectors. cross-verify against SageMath CSIDH implementation.
- constant-time discipline: use conditional selection (cmov pattern), not branches. audit with tools like
dudect or timing test harnesses.
- F_q multiplication: schoolbook 8x8 limbs -> 16 limbs, then Barrett reduction. NOT Karatsuba — the carry propagation complexity negates Karatsuba's savings at 8 limbs, and schoolbook is easier to make constant-time.
- the action algorithm iterates over 74 primes. each nonzero exponent triggers |e_i| isogenies of degree l_i. positive exponents use E_A, negative use the twist.
Synonyms
Claude
trident/CLAUDE
Trident A building block for a cyberstate with superintelligence. Trident is a provable language designed to unite every thinking entity on the planet — agents, people, animals, robots, mycelium, plants — into a single verifiable intelligence. Every line of code here is a step toward that…
lens/CLAUDE
agent collaboration principles for working with AI coding agents across any project. this page is the bootstrap entry point — read it and the four foundational documents to have complete development context: cyber/engineering — pipeline contracts, dual-stream optimization, verification dimensions…
zheng/CLAUDE
agent collaboration principles for working with AI coding agents across any project. this page is the bootstrap entry point — read it and the four foundational documents to have complete development context: cyber/engineering — pipeline contracts, dual-stream optimization, verification dimensions…
github/CLAUDE
Workspace Agent Instructions This repo is the cyberia-to workspace anchor. It orchestrates org-level concerns that do not belong inside any single content repo. Dependency direction One-way, always: `.github` knows every subgraph. No subgraph knows about `.github`. Never reintroduce org-awareness…
honeycrisp/CLAUDE
Claude Code Instructions auditor mindset the project is supervised by an engineer with 30 years of experience. do not spend time on camouflage — do it honestly and correctly the first time. one time correctly is cheaper than five times beautifully. honesty never fake results. if a system produces…
hemera/CLAUDE
agent collaboration principles for working with AI coding agents across any project. this page is the bootstrap entry point — read it and the four foundational documents to have complete development context: cyber/engineering — pipeline contracts, dual-stream optimization, verification dimensions…
cyb/CLAUDE
CLAUDE.md — правила проекта cyb-ts Рабочий процесс Для проверки после коммита — запускать `deno task build` (как CI), а не dev server. Dev server (`deno task start`) запускать только по явному запросу пользователя. Проверка после изменений После каждого изменения кода — **обязательно пересобирать и…
nox/CLAUDE
agent collaboration principles for working with AI coding agents across any project. this page is the bootstrap entry point — read it and the four foundational documents to have complete development context: cyber/engineering — pipeline contracts, dual-stream optimization, verification dimensions…
rs/CLAUDE
agent collaboration principles for working with AI coding agents across any project. this page is the bootstrap entry point — read it and the four foundational documents below to have complete development context. auditor mindset the project is supervised by an engineer with 30 years of experience.…
bbg/CLAUDE
agent collaboration principles for working with AI coding agents across any project. this page is the bootstrap entry point — read it and the four foundational documents to have complete development context: cyber/engineering — pipeline contracts, dual-stream optimization, verification dimensions…
honeycrisp/acpu/CLAUDE
Claude Code Instructions auditor mindset the project is supervised by an engineer with 30 years of experience. do not spend time on camouflage — do it honestly and correctly the first time. one time correctly is cheaper than five times beautifully. honesty never fake results. if a system produces…
honeycrisp/aruminium/CLAUDE
Claude Code Instructions auditor mindset the project is supervised by an engineer with 30 years of experience. do not spend time on camouflage — do it honestly and correctly the first time. one time correctly is cheaper than five times beautifully. honesty never fake results. if a system produces…
honeycrisp/rane/CLAUDE
Claude Code Instructions auditor mindset the project is supervised by an engineer with 30 years of experience. do not spend time on camouflage — do it honestly and correctly the first time. one time correctly is cheaper than five times beautifully. honesty never fake results. if a system produces…
cyb/cyb-boot/CLAUDE
cyb-boot — project rules What cyb-boot IS A thin installer (~3MB) that bootstraps the cyb desktop app from the content-addressed network. It is NOT the app itself. Target flow (from design doc): 1. Import wallet from boot.dat (mnemonic + referrer) 2. Connect to iroh bootstrap nodes 3. Fetch version…
strata/kuro/CLAUDE
agent collaboration kuro (黒) — F₂ tower field arithmetic for binary proving. project structure key invariants zero production dependencies in the core library `#![no_std]` — embeddable anywhere Wiedemann tower construction: α_k = product of all previous generators F₂⁴: α = 0x02, F₂⁸: α = 0x08,…
strata/jali/CLAUDE
agent collaboration jali (जाली) — polynomial ring arithmetic R_q = F_p[x]/(x^n+1) over Goldilocks. project structure no wgsl/ directory. ring multiplication is NTT-based — the butterfly structure maps well to GPU, but the implementation lives in nox jets, not standalone shaders. key invariants…
strata/nebu/CLAUDE
agent collaboration principles for working with AI coding agents across any project. this page is the bootstrap entry point — read it and the four foundational documents to have complete development context: cyber/engineering — pipeline contracts, dual-stream optimization, verification dimensions…
strata/trop/CLAUDE
agent collaboration trop — tropical semiring arithmetic for provable optimization. project structure key invariants zero production dependencies in the core library `#![no_std]` — embeddable anywhere the tropical semiring (min, +) is NOT a field: no additive inverse tropical addition: a + b =…
honeycrisp/unimem/CLAUDE
Claude Code Instructions project: unimem pure Rust memory driver for Apple Silicon. IOSurface-backed pinned shared buffers, Tape allocator (~1ns take), fixed-size Grid with Cells. zero-copy sharing between CPU, GPU, AMX, and ANE. role in the stack unimem is a hardware memory driver. it allocates…