agentlaunch

Command-line interface for AgentLaunch — one command to create, deploy, tokenize, and trade AI agents.

Install

# Global install (recommended for interactive use)
npm install -g agentlaunch

# Or run without installing
npx agentlaunch

Requires Node.js >= 18.

Configure

Get your API key from agentverse.ai/profile/api-keys, then:

agentlaunch config set-key YOUR_AGENTVERSE_API_KEY

Check your configuration:

agentlaunch config show

Override the API base URL (self-hosted or staging):

agentlaunch config set-url https://your-instance.example.com/api

Command Cheat Sheet

Every command at a glance. All support --json for machine-readable output.

Build

| Command | What It Does | |---------|-------------| | agentlaunch [name] | Create agent, deploy, open editor | | agentlaunch [name] --local | Scaffold only, no deploy | | agentlaunch scaffold <name> --type <t> | Generate agent from template | | agentlaunch init | Install toolkit into existing project | | agentlaunch deploy | Deploy agent.py to Agentverse | | agentlaunch optimize <addr> | Update README/description/avatar for ranking |

Tokenize & Trade

| Command | What It Does | |---------|-------------| | agentlaunch tokenize --agent <addr> | Create token + handoff link | | agentlaunch buy <addr> --amount <n> | Buy tokens on bonding curve | | agentlaunch sell <addr> --amount <n> | Sell tokens on bonding curve | | agentlaunch claim <wallet> | Claim 200 TFET + 0.001 tBNB (up to 3x) |

Wallet & Payments

| Command | What It Does | |---------|-------------| | agentlaunch wallet balances | Show FET + USDC + BNB balances | | agentlaunch wallet send <token> <to> <amt> | Transfer tokens | | agentlaunch wallet delegate <token> <amt> | Create spending approval link | | agentlaunch wallet allowance <token> | Check spending limit | | agentlaunch pay <to> <amt> --token <t> | Direct token payment |

Monitor

| Command | What It Does | |---------|-------------| | agentlaunch list | Browse tokens | | agentlaunch status <addr> | Check price/progress | | agentlaunch comments <addr> | List/post token comments | | agentlaunch holders <addr> | Token holder distribution |

Config

| Command | What It Does | |---------|-------------| | agentlaunch config set-key <key> | Store API key | | agentlaunch config show | Show current config | | agentlaunch config set-url <url> | Set custom API URL |

Commands

agentlaunch — the default command

Create an agent, deploy it to Agentverse, and open Claude Code — all in one step.

# Interactive — prompts for name, description, and API key
npx agentlaunch

# With name — skips the name prompt
npx agentlaunch my-agent

# Scaffold only (no deploy)
npx agentlaunch my-agent --local

# Machine-readable output for AI agents (no prompts, JSON only)
npx agentlaunch my-agent --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --local | Scaffold only, don't deploy to Agentverse | false (deploys by default) | | --template <t> | chat-memory (default), swarm-starter, custom, research, trading-bot, data-analyzer, price-monitor, gifter, consumer-commerce | chat-memory | | --json | Output only JSON (disables interactive prompts) | false |

Flow:

1. Prompts for name, description, and API key (if not already configured)
2. Scaffolds agent code from the chat-memory template (LLM + conversation memory)
3. Deploys to Agentverse (unless --local)
4. Opens Claude Code in the new project directory

Generated files:

• agent.py — ready-to-edit uAgents code with LLM integration and conversation memory
• README.md — quickstart instructions
• .env.example — required environment variables

agentlaunch scaffold <name>

Generate an agent project from any template without deploying.

# Default template (chat-memory)
agentlaunch scaffold my-bot

# Swarm-starter with commerce stack
agentlaunch scaffold my-service --type swarm-starter

# Swarm preset (writer, social, community, analytics, outreach, ads, strategy)
agentlaunch scaffold my-writer --type swarm-starter --preset writer

# Consumer commerce template
agentlaunch scaffold my-store --type consumer-commerce

Options:

| Flag | Description | Default | |------|-------------|---------| | --type <template> | Template: chat-memory, swarm-starter, custom, research, trading-bot, data-analyzer, price-monitor, gifter, consumer-commerce | chat-memory | | --preset <role> | Swarm preset role (only with swarm-starter) | — | | --json | Output only JSON | false |

agentlaunch deploy

Deploy agent.py to Agentverse. Auto-detects README.md and description from agentlaunch.config.json next to the agent file and uploads them as metadata to improve the agent's Agentverse ranking.

agentlaunch deploy

# Custom file and name
agentlaunch deploy --file ./my-agent.py --name "My Research Agent"

# JSON output
agentlaunch deploy --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --file <path> | Path to agent Python file | ./agent.py | | --name <name> | Display name on Agentverse | AgentLaunch Agent | | --json | Output only JSON | false |

Steps performed:

1. Create agent on Agentverse (with README + description if detected)
2. Upload code
3. Set AGENTVERSE_API_KEY and AGENTLAUNCH_API_KEY secrets
4. Start agent
5. Poll until compiled (up to 60 seconds)
6. Print optimization checklist (7 Agentverse ranking factors)

agentlaunch optimize <address>

Update metadata on an already-deployed agent to improve its Agentverse ranking. Auto-detects README.md in the current directory.

# Auto-detect README.md
agentlaunch optimize agent1q...

# Specify all fields
agentlaunch optimize agent1q... \
  --readme ./README.md \
  --description "AI research reports on demand" \
  --avatar https://example.com/avatar.png

# JSON output
agentlaunch optimize agent1q... --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --readme <path> | Path to README.md file | auto-detects ./README.md | | --description <text> | Short description (max 200 chars) | - | | --avatar <url> | Public URL for avatar image | - | | --json | Output only JSON | false |

agentlaunch tokenize

Create a token record on AgentLaunch and receive a handoff link for on-chain deployment.

agentlaunch tokenize \
  --agent agent1qXXXXXXXX \
  --name "My Research Agent" \
  --symbol RSRCH

# With optional fields
agentlaunch tokenize \
  --agent agent1qXXXXXXXX \
  --name "My Research Agent" \
  --symbol RSRCH \
  --description "Delivers on-demand research" \
  --chain 97 \
  --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --agent <address> | Agentverse address (agent1q...) | required | | --name <name> | Token name (max 32 chars) | required | | --symbol <symbol> | Ticker (2-11 chars) | required | | --description <desc> | Token description | - | | --image <url> | Token logo URL | - | | --chain <chainId> | 97 = BSC Testnet, 56 = BSC Mainnet | 97 | | --max-wallet <0\|1\|2> | Max wallet size: 0=unlimited, 1=0.5% (5M tokens), 2=1% (10M tokens) | 0 | | --initial-buy <amount> | FET to spend buying tokens immediately after deploy (0-1000 FET) | 0 | | --category <id> | Category ID for the token | 1 | | --json | Output only JSON | false |

The command returns a handoff link (e.g. https://agent-launch.ai/deploy/123). Share this with a human who connects their wallet and pays the deployment fee.

agentlaunch buy <address>

Buy tokens on the bonding curve. Requires WALLET_PRIVATE_KEY in .env.

# Buy with 10 FET
agentlaunch buy 0xAbCd... --amount 10

# Preview without executing
agentlaunch buy 0xAbCd... --amount 10 --dry-run

# Custom slippage
agentlaunch buy 0xAbCd... --amount 10 --slippage 3

# JSON output
agentlaunch buy 0xAbCd... --amount 10 --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --amount <fet> | FET amount to spend | required | | --dry-run | Preview only (no transaction) | false | | --slippage <pct> | Slippage tolerance (%) | 5 | | --chain <id> | Chain ID (97=testnet, 56=mainnet) | 97 | | --json | Output only JSON | false |

agentlaunch sell <address>

Sell tokens on the bonding curve for FET. Requires WALLET_PRIVATE_KEY in .env.

# Sell 50000 tokens
agentlaunch sell 0xAbCd... --amount 50000

# Preview
agentlaunch sell 0xAbCd... --amount 50000 --dry-run

# JSON output
agentlaunch sell 0xAbCd... --amount 50000 --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --amount <tokens> | Token amount to sell | required | | --dry-run | Preview only (no transaction) | false | | --chain <id> | Chain ID (97=testnet, 56=mainnet) | 97 | | --json | Output only JSON | false |

agentlaunch claim <wallet>

Claim testnet tokens (200 TFET + 0.001 tBNB). Up to 3 claims per wallet.

agentlaunch claim 0xMyWallet...

# JSON output
agentlaunch claim 0xMyWallet... --json

One claim covers the 120 TFET deploy fee with 80 TFET left for trading.

agentlaunch wallet

Multi-token wallet operations. Requires WALLET_PRIVATE_KEY in .env.

agentlaunch wallet balances

Show FET, USDC, and BNB balances.

agentlaunch wallet balances

# Specific chain
agentlaunch wallet balances --chain 56

# JSON output
agentlaunch wallet balances --json

agentlaunch wallet send <token> <to> <amount>

Transfer tokens to a recipient.

# Send 10 USDC
agentlaunch wallet send USDC 0xRecipient... 10

# Send FET
agentlaunch wallet send FET 0xRecipient... 50

agentlaunch wallet delegate <token> <amount>

Generate a handoff link for a human to approve a spending limit.

agentlaunch wallet delegate FET 100 --spender 0xAgent...

agentlaunch wallet allowance <token>

Check the spending limit an owner has approved for a spender.

agentlaunch wallet allowance FET --owner 0xOwner... --spender 0xAgent...

agentlaunch pay <to> <amount>

Direct token payment. Requires WALLET_PRIVATE_KEY in .env.

# Pay in FET (default)
agentlaunch pay 0xRecipient... 10

# Pay in USDC
agentlaunch pay 0xRecipient... 10 --token USDC

# JSON output
agentlaunch pay 0xRecipient... 10 --token USDC --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --token <symbol> | Token to pay with (FET, USDC) | FET | | --chain <id> | Chain ID | 97 | | --json | Output only JSON | false |

agentlaunch init

Install the AgentLaunch toolkit into an existing project. Adds SDK dependency and creates starter files.

cd my-existing-project
agentlaunch init

agentlaunch list

List tokens on AgentLaunch.

agentlaunch list

# Custom options
agentlaunch list --limit 20 --sort market_cap

# JSON output for AI agents
agentlaunch list --limit 5 --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --limit <n> | Number of tokens to show (1-100) | 10 | | --sort <by> | trending, latest, market_cap | latest | | --json | Output only JSON | false |

agentlaunch status <address>

Show detailed status of a token by its contract address.

agentlaunch status 0xAbCdEf...

# JSON output
agentlaunch status 0xAbCdEf... --json

agentlaunch comments <address>

List or post comments on a token.

# List comments for a token
agentlaunch comments 0xAbCdEf...

# Post a comment (requires API key configured)
agentlaunch comments 0xAbCdEf... --post "Great agent, just bought some!"

# JSON output for machine parsing
agentlaunch comments 0xAbCdEf... --json

Options:

| Flag | Description | Default | |------|-------------|---------| | --post <message> | Post a new comment (max 500 chars). Requires API key. | - | | --json | Output only JSON | false |

agentlaunch holders <address>

Show the token holder distribution table.

agentlaunch holders 0xAbCdEf...

# JSON output
agentlaunch holders 0xAbCdEf... --json

agentlaunch config

Manage CLI configuration.

agentlaunch config set-key <apiKey>   # Store your API key
agentlaunch config show               # Show current config
agentlaunch config set-url <url>      # Set custom API base URL

Common Workflows

Build and deploy a new agent

npx agentlaunch my-bot
# -> Scaffolds, deploys, opens editor
# Edit agent.py, then:
agentlaunch optimize agent1q...
agentlaunch tokenize --agent agent1q... --name "MyBot" --symbol MBOT
# Open the handoff link, connect wallet, sign

Deploy a swarm

agentlaunch scaffold writer --type swarm-starter --preset writer
agentlaunch scaffold social --type swarm-starter --preset social
agentlaunch deploy --file writer/agent.py --name "Writer Agent"
agentlaunch deploy --file social/agent.py --name "Social Agent"
agentlaunch optimize agent1qWriter...
agentlaunch optimize agent1qSocial...

Trade tokens

# Check what's available
agentlaunch list --sort trending

# Preview a buy
agentlaunch buy 0xToken... --amount 10 --dry-run

# Execute
agentlaunch buy 0xToken... --amount 10

# Check holdings
agentlaunch wallet balances

# Sell
agentlaunch sell 0xToken... --amount 50000

Multi-token payments

# Check balances
agentlaunch wallet balances

# Pay in USDC
agentlaunch pay 0xRecipient... 10 --token USDC

# Set up delegation
agentlaunch wallet delegate FET 100 --spender 0xAgent...

# Check delegation
agentlaunch wallet allowance FET --owner 0xMe... --spender 0xAgent...

JSON Mode (Machine-Readable for AI Agents)

Every command supports --json. In JSON mode:

• All output is valid JSON to stdout
• No decorative text, tables, or prompts
• Interactive prompts are disabled (missing required flags cause an error JSON)
• Exit code 0 on success, 1 on error

Example — AI agent creates and deploys programmatically:

RESULT=$(npx agentlaunch my-analyst --json)

ADDRESS=$(echo "$RESULT" | python3 -c "import sys,json; print(json.load(sys.stdin)['agentAddress'])")
echo "Agent deployed: $ADDRESS"

Templates

| Template | Use Case | |----------|----------| | chat-memory | LLM + conversation memory (default) — smart conversations out of the box | | swarm-starter | Full commerce stack — agents that charge for services | | consumer-commerce | Multi-token payments, invoices, fiat onramp | | custom | Blank slate — add your own business logic | | price-monitor | Watch token prices, send alerts | | trading-bot | Buy/sell signal generation | | data-analyzer | On-chain data analysis | | research | On-demand research reports and analysis | | gifter | Treasury wallet, reward distribution |

Platform Constants

These values are enforced by the deployed smart contracts and cannot be changed by the CLI or platform:

| Constant | Value | Notes | |----------|-------|-------| | Token deployment fee | 120 FET | Read dynamically from contract; may change via multi-sig governance | | Graduation target | 30,000 FET | Auto DEX listing when reached | | Total buy supply | 800,000,000 tokens | Fixed bonding curve supply | | Trading fee | 2% | 100% to protocol treasury. There is NO creator fee. |

Cross-References

• SDK: agentlaunch-sdk — TypeScript client this CLI wraps
• MCP Server: agent-launch-mcp — Same operations as Claude Code tools
• Templates: agentlaunch-templates — Agent blueprints

Resources

• AgentLaunch Platform
• API Documentation
• Agent Integration Guide
• skill.md — Machine-readable capability discovery
• Agentverse
• Agentverse API Keys