# dstack-cli
Development stack CLI for AI-assisted multi-repo work.
## Install
```bash
cargo install dstack-cli
```
## Commands
| `dstack memory load` | Load memory fields for a project |
| `dstack memory save <key> <value>` | Save a key-value pair to memory |
| `dstack memory query <pattern>` | Search memory by keyword |
| `dstack memory export` | Export all memory as JSON |
| `dstack deploy <service>` | Deploy a service (build + restart + smoke test) |
| `dstack deploy --all` | Deploy all configured services |
| `dstack sync` | Git sync across tracked repos |
| `dstack sync --status` | Show sync status without pushing |
| `dstack skills list` | List available skills |
| `dstack skills sync` | Install all skills |
| `dstack audit` | Quality audit summary |
| `dstack audit --pre-commit` | Pre-commit quality gate |
| `dstack init <dir>` | Scaffold a multi-platform plugin |
## Configuration
Create `~/.config/dstack/config.toml`:
```toml
[memory]
backend = "file" # or "eruka"
[repos]
tracked = ["~/projects/my-repo"]
[deploy.my-service]
build = "cargo build --release"
service = "my-service"
smoke = "curl -sf http://localhost:3000/health"
[deploy.my-containers]
deploy_type = "docker-compose"
service = "my-service"
compose_file = "~/docker-compose.yml"
smoke = "curl -sf http://localhost:8080/health"
```
## Plugin
dstack includes a multi-platform plugin for AI coding agents. See the [plugin README](../../plugin/README.md).
## License
MIT