Skip to content

Operations

This section is for the person running Myelin: installing it, running it, and proving it works against a real (or mock) CKB node.

Pages

  • CLI reference

    Every myelin-cli subcommand, what it takes, what it produces.

  • Production gate

    The scripts/myelin_production_gate.sh end-to-end script: what it runs, in what order, what each step proves.

  • Local CKB devnet smoke

    The scripts/myelin_ckb_devnet_smoke.sh script: live carrier submissions, type-script execution, and live rejection of tampered carriers on a parent CKB devnet.

What "running Myelin" looks like today

There is no daemon. There is no P2P layer. There is no RPC server. The way you run Myelin today is:

  1. Compile the CLI.
  2. Run subcommands that produce and verify reports on disk.
  3. Run scripts that orchestrate the subcommands and assert properties of the reports.

If you want a daemon or RPC server, that's a separate piece of work — Myelin today is a kernel + CLI, not a service.

What "production evidence" looks like today

Two scripts in scripts/ produce the evidence:

scripts/myelin_production_gate.sh       # full local gate
scripts/myelin_teeworlds_acceptance.sh  # narrower Teeworlds gate
scripts/myelin_ckb_devnet_smoke.sh      # live CKB devnet path

Each one leaves behind a set of JSON reports in reports/ that record what was measured, what was projected, and what was submitted. Those reports are the evidence; the CLI exit codes are the gate.

Where to go next