skills/aglais-xqvm-quantum-vm/SKILL.md
Expertise in Aglais XQVM, a hardware-agnostic Rust quantum virtual machine for QUBO/Ising binary optimization models targeting quantum annealers.
npx skillsauth add aradotso/trending-skills aglais-xqvm-quantum-vmInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Skill by ara.so — Daily 2026 Skills collection.
Aglais XQVM is a hardware-agnostic virtual machine for quantum computing written in Rust. It provides a unified bytecode intermediate representation for binary optimization problems (QUBO/Ising formulations) targeting quantum annealers — think LLVM for quantum computing. The VM is stack-based with a 256-slot register file, supports no_std + alloc for WASM/bare-metal deployment, and ships four crates: bytecode, assembler, disassembler, and interpreter.
# Install Rust stable
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
# Install dev tools (cargo-nextest, clippy, etc.)
make deps
git clone https://github.com/QuipNetwork/xq-rs
cd xq-rs
cargo build --release
# Binaries: target/release/xqasm, target/release/xqdism, target/release/xqvm
# Cargo.toml
[dependencies]
aglais-xqvm-bytecode = { path = "crates/bytecode" }
aglais-xqvm-vm = { path = "crates/vm" }
For no_std environments (WASM, bare-metal):
[dependencies]
aglais-xqvm-bytecode = { path = "crates/bytecode", default-features = false, features = ["alloc"] }
| Crate | Binary | Role |
|---|---|---|
| aglais-xqvm-bytecode | — | Opcode table, instruction types, builder, binary codec, stream reader |
| aglais-xqvm-asm | xqasm | Text assembler: .xqasm → .xqbc bytecode |
| aglais-xqvm-disasm | xqdism | Bytecode → human-readable listing |
| aglais-xqvm-vm | xqvm | Bytecode interpreter: stack, registers, QUBO/Ising execution |
xqasm — Assembler# Assemble a source file to bytecode
xqasm program.xqasm -o program.xqbc
# Assemble with verbose output
xqasm program.xqasm -o program.xqbc --verbose
xqdism — Disassembler# Inspect bytecode encoding as human-readable listing
xqdism program.xqbc
# Pipe to file
xqdism program.xqbc > listing.txt
xqvm — Interpreter# Execute bytecode
xqvm program.xqbc
# Run with debug output (if supported)
xqvm program.xqbc --debug
xqasm problem.xqasm -o problem.xqbc && xqdism problem.xqbc && xqvm problem.xqbc
The assembler accepts .xqasm text files. The VM is stack-based; most instructions pop operands from the stack and push results.
; push two integers and add them
PUSH 10
PUSH 32
ADD
HALT
PUSH 42
STORE 0 ; pop stack → register 0
LOAD 0 ; push register 0 → stack
PUSH 10
PUSH 3
ADD ; stack: [13]
PUSH 7
SUB ; stack: [6]
PUSH 2
MUL ; stack: [12]
PUSH 4
DIV ; stack: [3]
; build a 3-element vector [1, 2, 3]
PUSH 1
PUSH 2
PUSH 3
PUSH 3 ; length
VEC ; stack: [Vec([1,2,3])]
STORE 1
; XQMX_NEW n creates an n-variable QUBO model
PUSH 4
XQMX_NEW ; stack: [XqmxModel(4 vars)]
STORE 2
; set quadratic coupling Q[i][j] = weight
LOAD 2
PUSH 0 ; i
PUSH 1 ; j
PUSH -1 ; weight (integer encoding)
XQMX_SET_Q ; modifies model in reg 2
; set linear bias h[i] = weight
LOAD 2
PUSH 0
PUSH 5
XQMX_SET_H
; evaluate energy of a candidate solution
LOAD 2 ; model
PUSH 0 ; sample register (XqmxSample)
XQMX_EVAL ; pushes energy onto stack
; RANGE lo hi → loop stack entry, ITER steps through it
PUSH 0
PUSH 5
RANGE ; loop i in 0..5
ITER ; advance; jumps past matching END_ITER when done
LOAD 0
PUSH 1
ADD
STORE 0
END_ITER
HALT
PUSH 0
loop:
PUSH 1
ADD
DUP
PUSH 10
LT
JMP_TRUE loop
HALT
Use aglais-xqvm-bytecode to construct programs programmatically:
use aglais_xqvm_bytecode::{BytecodeBuilder, Instruction, Opcode};
fn build_add_program() -> Vec<u8> {
let mut builder = BytecodeBuilder::new();
builder.emit(Instruction::Push(10));
builder.emit(Instruction::Push(32));
builder.emit(Instruction::Add);
builder.emit(Instruction::Halt);
builder.finish()
}
use aglais_xqvm_bytecode::StreamReader;
fn decode(bytes: &[u8]) {
let mut reader = StreamReader::new(bytes);
while let Some(instr) = reader.next_instruction().unwrap() {
println!("{:?}", instr);
}
}
use aglais_xqvm_vm::Vm;
fn main() {
// Load bytecode from a file
let bytecode = std::fs::read("program.xqbc").expect("read bytecode");
let mut vm = Vm::new();
vm.load(&bytecode).expect("load");
vm.run().expect("run");
// Inspect top of stack after execution
if let Some(val) = vm.stack_top() {
println!("Result: {:?}", val);
}
}
use aglais_xqvm_vm::{Vm, Value};
fn run_and_inspect(bytecode: &[u8]) -> Value {
let mut vm = Vm::new();
vm.load(bytecode).unwrap();
vm.run().unwrap();
vm.register(0).cloned().unwrap_or(Value::Int(0))
}
The crates/vm/examples/tsp/ directory contains a complete Travelling Salesman Problem encoded as a QUBO driven by a Rust harness. The pattern is:
.xqasm files parameterised by those coefficients.xqasm / xqvm.// crates/vm/examples/tsp/main.rs pattern
use std::process::Command;
fn assemble_and_run(src: &str, out: &str) {
let asm = Command::new("xqasm")
.args([src, "-o", out])
.status()
.expect("xqasm failed");
assert!(asm.success());
let run = Command::new("xqvm")
.arg(out)
.status()
.expect("xqvm failed");
assert!(run.success());
}
fn main() {
assemble_and_run("init.xqasm", "init.xqbc");
assemble_and_run("problem.xqasm", "problem.xqbc");
assemble_and_run("eval.xqasm", "eval.xqbc");
}
; 2-variable QUBO: minimise x0 - x1 + 2*x0*x1
PUSH 2
XQMX_NEW
STORE 0
LOAD 0
PUSH 0
PUSH -1 ; h[0] = -1 (linear)
XQMX_SET_H
LOAD 0
PUSH 1
PUSH -1 ; h[1] = -1 (linear)
XQMX_SET_H
LOAD 0
PUSH 0
PUSH 1
PUSH 2 ; Q[0][1] = 2 (quadratic)
XQMX_SET_Q
HALT
PUSH 4
XQMX_NEW
STORE 0
PUSH 0
PUSH 4
RANGE
ITER
; register 1 holds current loop index after ITER
LOAD 0
LOAD 1 ; index i
LOAD 1 ; index i (diagonal → linear term)
PUSH -1
XQMX_SET_Q
END_ITER
HALT
#![no_std]
extern crate alloc;
use alloc::vec::Vec;
use aglais_xqvm_bytecode::StreamReader;
pub fn decode_instructions(bytes: &[u8]) -> Vec<alloc::string::String> {
let mut reader = StreamReader::new(bytes);
let mut out = Vec::new();
while let Ok(Some(instr)) = reader.next_instruction() {
out.push(alloc::format!("{:?}", instr));
}
out
}
# Run all lints and tests (mirrors CI)
make all
# Run only tests
cargo test --workspace
# Run lints
cargo clippy --workspace --all-targets -- -D warnings
# Format
cargo fmt --all
# Run a specific example
cargo run --example tsp --manifest-path crates/vm/Cargo.toml
The opcode table in crates/bytecode/src/types/table.rs is the single source of truth for all 76 instructions. Key categories:
| Category | Instructions |
|---|---|
| Stack | PUSH, POP, DUP, SWAP |
| Registers | LOAD, STORE |
| Arithmetic | ADD, SUB, MUL, DIV, NEG |
| Comparison | EQ, LT, GT, LE, GE |
| Control flow | JMP, JMP_TRUE, JMP_FALSE, CALL, RET, HALT |
| Iteration | RANGE, ITER, END_ITER |
| Vectors | VEC, VEC_GET, VEC_SET, VEC_LEN |
| QUBO/Ising | XQMX_NEW, XQMX_SET_Q, XQMX_SET_H, XQMX_EVAL, XQMX_SAMPLE |
All operands are big-endian. The binary format is a bare instruction stream with no file header.
xqasm: command not foundEnsure target/release is on $PATH or use the full path:
export PATH="$PWD/target/release:$PATH"
The VM is strictly stack-based. Every instruction that pops values requires them to be present. Check that PUSH / LOAD precedes every operation, and that loops don't consume values without restoring the stack balance.
ITER never terminatesRANGE pushes loop bounds onto the loop stack (separate from the value stack). Ensure every RANGE has a matching END_ITER and that the range bounds (lo, hi) are pushed in the correct order (lo first, hi second).
no_std environmentDisable default features and enable the alloc feature on aglais-xqvm-bytecode:
aglais-xqvm-bytecode = { ..., default-features = false, features = ["alloc"] }
The VM crate (aglais-xqvm-vm) requires std and is not suitable for bare-metal.
Use xqdism to verify the assembler output before running:
xqasm suspect.xqasm -o suspect.xqbc
xqdism suspect.xqbc # check instruction sequence and operand values
xqvm suspect.xqbc
AGPL-3.0-or-later. Embedding in proprietary network services requires source disclosure under the AGPL.
development
```markdown --- name: compose-performance-skills description: Install and use the skydoves/compose-performance-skills agent skill library to diagnose and fix Jetpack Compose performance issues including stability, recomposition, lazy layouts, modifiers, side effects, and build configuration. triggers: - "my composable recomposes too often" - "LazyColumn drops frames during scroll" - "diagnose Compose stability issues" - "fix unnecessary recomposition in Jetpack Compose" - "optimize Com
development
Headless iOS Simulator manager with host-side HID input injection, 60fps streaming, and device farm web UI for iOS 26
development
```markdown --- name: claude-code-game-studios description: Turn Claude Code into a full 49-agent game dev studio with 72 workflow skills, automated hooks, and a real studio hierarchy for Godot, Unity, and Unreal projects. triggers: - "set up claude code game studios" - "use ai agents for game development" - "set up game dev studio with claude" - "add game studio agents to my project" - "how do I use claude code for game dev" - "set up godot unity unreal ai workflow" - "49 agents g
development
```markdown --- name: xq-py-quantum-vm description: Python implementation of the Quip Network's quantum virtual machine (xqvm) triggers: - quantum virtual machine python - xqvm quip network - quantum circuit simulation python - xq-py quantum vm - quip network quantum python - simulate quantum gates python - quantum vm xqvm - xqvm-py quantum circuit --- # xq-py Quantum Virtual Machine > Skill by [ara.so](https://ara.so) — Daily 2026 Skills collection. `xqvm-py` is a Python impl