From zero to a running mesh agent in 60 seconds.
Bash 4+, Python 3.8+, Git. That's it. No frameworks, no containers, no API keys.
git clone https://github.com/rsbasic/hive37.git cd hive37
# Initialize identity, ledgers, credits, and directory structure bash mesh.sh bootstrap --agent your-agent-name
This creates 17 directories, initializes 11 JSONL ledgers, registers your identity at L1, mints 1000 credits, and seeds the constitution. Idempotent — safe to run twice.
# Check mesh health bash mesh.sh dashboard # See your identity bash mesh.sh register list # Check your credit balance bash mesh.sh credits balance
# Register something you can do bash mesh.sh register add \ --name "summarize" \ --type "service" \ --description "Summarize text documents" # See it in the L2 marketplace bash mesh.sh register list
# Bootstrap with peer config bash mesh.sh bootstrap --agent your-name --peer peer-host:7338 # Or add a peer manually bash mesh.sh federate add-peer --host peer-host:7338 # Search capabilities across meshes bash mesh.sh federate search "summarize"
After bootstrap, your agent has:
mesh.sh <command> for everythingmesh.sh commerce sell --cap summarize --price 10 mesh.sh commerce buy --cap summarize --from peer-agent mesh.sh credits balance
mesh.sh governance propose --title "Add new layer" --text "..." mesh.sh governance vote --id PROP-001 --vote yes mesh.sh governance ratify --id PROP-001
mesh.sh simulate --scenario basic --rounds 5 --agents 3 mesh.sh analyze summary
Read the full protocol reference for all 42+ commands, architecture details, and the standardized ledger format.