Prerequisites -- Axon

September 5, 2026 · View on GitHub

Required tools and versions before developing or deploying.

Required tools

Tool	Version	Install	Purpose
Rust	1.97.1	rustup.rs	Compiler (pinned via `rust-toolchain.toml`)
Docker	24+	docs.docker.com	Infrastructure services
Docker Compose	v2+	Bundled with Docker	Service orchestration
just	latest	`cargo install just`	Task runner
Node.js	24+	nodejs.org	Web UI runtime and npm
jq	1.6+	System package manager	JSON parsing in scripts
curl	any	System package manager	HTTP testing
Python	3.10+	System package manager	Scripts and analysis tools

Verify

rustc --version          # rustc 1.97.1
cargo --version          # cargo 1.97.1
docker --version         # Docker 24+
docker compose version   # Docker Compose v2+
just --version           # just x.y.z
node --version           # v24.0.0 or newer
npm --version            # npm bundled with Node.js
jq --version             # jq-1.6+
python3 --version        # Python 3.10+

Automated setup

The setup script installs everything:

just setup
# or directly:
./scripts/dev-setup.sh

This installs: Rust toolchain components and cargo tools, then runs npm ci for the web panel and verifies all prerequisites. If just is not yet installed, dev-setup.sh installs it.

Optional tools (recommended)

Tool	Install	Purpose
kache	`mise install`	Compilation cache (global `rustc-wrapper`)
mold	System package manager	Fast linker (auto-detected by Justfile)
cargo-nextest	`cargo install cargo-nextest`	Faster parallel test runner
cargo-deny	`cargo install cargo-deny`	Dependency auditing
cargo-llvm-cov	`cargo install cargo-llvm-cov`	Code coverage
lefthook	`cargo install lefthook`	Git hooks
gh	cli.github.com	GitHub CLI for PRs and issues

Infrastructure services

Infrastructure runs via Docker Compose. No manual installation needed:

Service	Docker image	Host port	Purpose
Qdrant	qdrant/qdrant:v1.18.2	53333	Vector store
TEI	ghcr.io/huggingface/text-embeddings-inference:89-1.9	52000	Embedding generation
Chrome	Custom (`config/chrome/`)	6000	Headless browser

Start all infrastructure:

just services-up

GPU requirements (optional)

TEI benefits from GPU acceleration for embedding generation:

NVIDIA GPU with CUDA drivers
docker compose --env-file ~/.axon/.env -f docker-compose.prod.yaml up -d
CPU-only: use an external TEI_URL or override the TEI image/settings to remove the NVIDIA device reservation before starting the default services compose file.

System resources

Recommended minimums for local development:

Resource	Minimum	Recommended
RAM	8 GB	16+ GB
Disk	20 GB	50+ GB (for Qdrant data and TEI models)
CPU	4 cores	8+ cores (concurrency scales with cores)

Quick start

git clone https://github.com/dinglebear-ai/axon.git ~/workspace/axon
cd ~/workspace/axon
just setup               # Install all tools
mkdir -m 700 -p ~/.axon
cp .env.example ~/.axon/.env && chmod 600 ~/.axon/.env
# Edit ~/.axon/.env with credentials
just services-up          # Start infrastructure
just dev                  # Build + run full stack