# Xian Node > Operate current Xian nodes with xian-tech-cli and xian-stack. Use when joining canonical networks, creating private networks, running validators or service nodes, and inspecting runtime health. Use when this capability is needed. - Skill: `tomevault-io/xian-node` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/xian-node` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/xian-node/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/xian-node --- # Xian Node Skill Use this skill for the current Xian operator flow. Important packaging detail: - install package: `xian-tech-cli` - command name: `xian` ## Installation ```bash pip install xian-tech-cli ``` ## Core Model Xian node operations are split across: - `xian-cli` for operator workflows and node profiles - `xian-stack` for the runtime backend and Docker services - canonical network manifests for pinned release images and network metadata Default production posture: - canonical networks use manifest-pinned registry images - development/private networks can opt into `local_build` ## Quick Reference ```bash # Join a canonical network uv run xian network join validator-1 --network mainnet --moniker "validator-1" # Start and inspect uv run xian node start validator-1 uv run xian node status validator-1 uv run xian node endpoints validator-1 uv run xian node health validator-1 # Stop uv run xian node stop validator-1 ``` ## Join Mainnet or Testnet ### Validator Example ```bash uv run xian network join validator-1 \ --network mainnet \ --moniker "validator-1" uv run xian node start validator-1 uv run xian node status validator-1 uv run xian node endpoints validator-1 ``` ### Service Node Example ```bash uv run xian network join service-1 \ --network mainnet \ --moniker "service-1" \ --service-node \ --enable-dashboard \ --enable-monitoring uv run xian node start service-1 uv run xian node health service-1 ``` ## Local Build Override Use this when you explicitly want the local checked-out repos instead of the published canonical image. ```bash uv run xian network join dev-validator \ --network mainnet \ --moniker "dev-validator" \ --node-image-mode local_build ``` ## New Network / Private Network ```bash uv run xian network create localnet-1 \ --chain-id xian-localnet-1 \ --moniker "bootstrap-1" \ --force uv run xian node start localnet-1 ``` Use `xian network create` when you need a new private network definition rather than joining a canonical one. ## Optional Runtime Layers The node profile can enable: - `--service-node` for BDS/indexed reads - `--enable-dashboard` for the explorer/dashboard service - `--enable-monitoring` for Prometheus and Grafana - `--enable-intentkit` for the optional `xian-intentkit` sidecar stack Example: ```bash uv run xian network join agent-node \ --network mainnet \ --service-node \ --enable-dashboard \ --enable-monitoring \ --enable-intentkit ``` ## What `node status` Should Tell You The current status flow is richer than a simple process check. It should be the first command you reach for. ```bash uv run xian node status validator-1 ``` Expect it to surface: - profile/network identity - runtime/backend reachability - current image mode and image references - manifest-derived release provenance when present - dashboard / monitoring / intentkit reachability when enabled ## Endpoints and Health ```bash uv run xian node endpoints validator-1 uv run xian node health validator-1 ``` Use: - `node endpoints` for local URLs - `node health` for machine-readable checks - `node status` for the operator summary ## Direct Stack Backend Use the raw backend only when you are working directly in `xian-stack` and need to bypass the higher-level CLI. ```bash python3 ./scripts/backend.py start --no-service-node --no-dashboard --no-monitoring python3 ./scripts/backend.py status --no-service-node --no-dashboard --no-monitoring python3 ./scripts/backend.py endpoints --no-service-node --no-dashboard --no-monitoring python3 ./scripts/backend.py health --no-service-node --no-dashboard --no-monitoring python3 ./scripts/backend.py stop --no-service-node --no-dashboard --no-monitoring ``` ## Operational Guidance - Prefer canonical manifest-backed images for real networks. - Use `local_build` only when you intentionally need unreleased local code. - Use service-node mode when you need BDS, GraphQL, explorers, or indexed bot reads. - Use the dashboard for human inspection, not as the only source of truth. - Use `node health` and logs for automated checks and incident triage. ## Resources - [xian-tech-cli on PyPI](https://pypi.org/project/xian-tech-cli/) - [xian-technology/xian-cli](https://github.com/xian-technology/xian-cli) - [xian-technology/xian-stack](https://github.com/xian-technology/xian-stack) - [xian node docs](https://docs.xian.org/node/) --- > Converted and distributed by [TomeVault](https://tomevault.io/claim/xian-technology) — claim your Tome and manage your conversions.