| name | eth2-quickstart |
| description | Use this skill whenever the task involves an Ethereum validator node — setting up a node, installing Geth/Besu/Nethermind/Reth or Prysm/Lighthouse/Teku/other clients, configuring MEV-Boost or Commit-Boost, checking node health, listing active validators, managing validator exits, BLS-to-execution withdrawal changes, or 0x02 compounding creation, operating services, cleaning node data, or updating clients. Routes all actions through `./scripts/eth2qs.sh` with strict safety guardrails. If in doubt about an Ethereum node task, trigger this skill. |
| metadata | {"openclaw":{"skillKey":"eth2-quickstart"}} |
Eth2 Quickstart
Use this skill for Ethereum node workflows inside an eth2-quickstart checkout. Publish/install it through ClawHub or npx clawhub, then run it from the repo so it can use the canonical wrapper commands.
Routing
- For end-to-end node operator goals, read operator.md.
- Install/bootstrap: use
./scripts/eth2qs.sh bootstrap .... Read workflow.md.
- Configure or rerun phases: use
./scripts/eth2qs.sh configure, phase1, phase2, or monad-install. Read workflow.md.
- Diagnose or inspect status: use
./scripts/eth2qs.sh doctor --json for health gates, ./scripts/eth2qs.sh stats --json for machine-readable triage, ./scripts/eth2qs.sh debug --json --service <name> for structured RCA, ./scripts/eth2qs.sh update-check --json for freshness/drift, and ./scripts/eth2qs.sh repair for bounded repair preview/apply. Read outputs.md.
- For validator lifecycle operations, use
./scripts/eth2qs.sh validators --json to list active validators and ./scripts/eth2qs.sh validator-exit, ./scripts/eth2qs.sh validator-withdrawal-changes, ./scripts/eth2qs.sh validator-create-0x02, or ./scripts/eth2qs.sh validator-manage for focused exit / withdrawal-change / compounding / consolidation flows. Rehearse withdrawal changes with ./scripts/eth2qs.sh validator-withdrawal-changes --dry-run --generate --submit --yes before touching live keys. Read commands.md and operator.md.
- Operate services, logs, cleanup, or updates: use the mapped commands in commands.md.
- Before destructive or privilege-sensitive actions, load safety.md.
- For server recommendation or host-fit questions, read sizing.md.
- For concrete user-task phrasing, read examples.md.
- For Claude Code / Codex native tool use, read mcp.md.
- For durable repo-scoped improvement loops, read improvement.md.
Rules
- Do not invent new command entrypoints when an existing wrapper command fits.
- Prefer
./scripts/eth2qs.sh over calling utility scripts directly.
- Do not generate validator keys.
- Do not remove secrets.
- Require human confirmation before destructive cleanup, host-wide changes, or steps that require reboot/root.
- Treat
doctor --json as the canonical health gate, stats --json as the canonical machine-readable triage surface, and monitor export --json as the canonical compact alert/dashboard surface.
- Treat
validators --json as the canonical read-only validator inventory surface; use it to inspect local validator index/status/balance before taking exit, withdrawal-change, or compounding actions. The JSON output also carries inventory freshness metadata and beacon query status.
References
- workflow.md: choose bootstrap vs configure vs phase workflows
- operator.md: operator-focused install, resume, run, update, and cleanup flows
- commands.md: canonical command mapping
- safety.md: destructive boundaries and secrets policy
- sizing.md: server sizing and host-fit guidance
- outputs.md: expected outputs and checks for agents
- examples.md: concrete prompt/task examples for common operator goals
- mcp.md: Claude Code / Codex MCP server usage and safety contract
- improvement.md: how to turn failures and test output into durable repo learnings
