@tenseuronp/cli

Tenseuron CLI

Mine or validate AI tasks on the Tenseuron Protocol from any machine — cloud APIs or local LLMs.

Quick start

# 1. Set up your miner (interactive)
npx @tenseuronp/cli miner setup

# 2. Run it
npx @tenseuronp/cli miner

# Or set up a validator
npx @tenseuronp/cli validator setup
npx @tenseuronp/cli validator

# Check what's running
npx @tenseuronp/cli status

No git clone. No cd. Works from any directory.

Providers

Cloud APIs — pay-as-you-go, low setup:

• DeepSeek (cheap, fast)
• OpenAI (gpt-4o-mini, etc.)
• Anthropic (Claude Haiku)
• Groq (Llama on Groq)
• Mistral

Local LLMs — your own GPU, zero per-token cost:

• Ollama (http://localhost:11434/v1)
• LM Studio (http://localhost:1234/v1)
• Any OpenAI-compatible endpoint

Local providers don't need an API key.

How it works

• Miner: polls open tasks on the network → runs them through your AI → submits the result → earns network tokens when validators score it.
• Validator: polls miner submissions → scores them 0–100 with your AI → earns network tokens for accurate scoring.
• Payouts: 50% miner / 30% validators / 10% network creator / 10% launchpad treasury, per completed task.

Config

Setup writes to ~/.tenseuron/miner.env or ~/.tenseuron/validator.env. The interactive wizard (tenseuron miner setup) fills in the placeholders below for you — edit by hand only if you need to change something later.

MINER_ADDRESS=<your_solana_wallet_pubkey>
NETWORK_ID=<network_id_from_the_launchpad>
PROVIDER=custom              # custom = local OpenAI-compatible
MODEL=llama3.2:1b
API_KEY=local
CUSTOM_ENDPOINT=http://localhost:11434/v1
POLL_INTERVAL=5000
BACKEND_URL=https://api.tenseuron.com

Requirements

• Node.js 18+
• A Solana wallet (Phantom recommended) — you only paste its public address; the CLI never sees your private key
• For local LLM mode: Ollama or LM Studio running locally

Troubleshooting

• "You have not joined this network" — Re-run setup; the join step is automatic.
• "AI failed: ECONNREFUSED localhost:11434" — Ollama isn't running. Start it: ollama serve.
• "unable to allocate CPU buffer" — Local model too big for your RAM. Try a smaller model: ollama pull llama3.2:1b.
• Validator console shows only dots — Healthy idle. No submissions pending. Send a task from the WalletAI extension to test.

License

MIT