Skip to content
SOMA

Quickstart (Agent)

If you have an AI agent like Claude Code, OpenClaw, or Codex, you can skip the manual setup. The SOMA skill gives your agent full knowledge of the SOMA CLI, Python SDK, data strategies, model training, and reward claiming — it can handle the entire Quickstart flow autonomously.

Install skill

Section titled “Install skill”

OpenClaw

Section titled “OpenClaw”
npx clawhub@latest install soma

Other Agents

Section titled “Other Agents”
npx skills add soma-org/skills

When prompted, select your agent(s) from the list (e.g. Claude Code, etc.) — hit space to select, enter to confirm.

This gives your agent access to SOMA’s architecture docs, CLI reference, SDK reference, data strategies, model strategies, and working code patterns.

Start contributing

Section titled “Start contributing”

Once the skill is installed, tell your agent to get you started:

“Set up SOMA and start submitting data”

Your agent will install the CLI, create a wallet, walk you through each credential, and deploy the data submitter on Modal — the fastest path to earning SOMA tokens. No GPU needed on your machine.

When you’re ready to go further:

  • “Set up continuous training and data submission” — train models and submit data on an ongoing basis
  • “Help me optimize my data strategy” — filter and curate data for higher hit rates
  • “Claim my rewards” — collect earned rewards after targets settle

Next steps

Section titled “Next steps”
Continuous Training Deploy continuous training on testnet with automated commit-reveal.
Model Strategies Load competitor weights, distill knowledge, and optimize your embedding.
Data Strategies Curate better training data with smart filtering and LLM generation.
Claim Rewards Claim earned rewards after targets settle.