Skip to content

Solizardking/clawd-bot / agent/skills/solana-clawd

solana-clawd

Verified

One-shot setup and operation guide for the solana-clawd agentic engine. Use when: cloning the repo, setting up MCP tools, starting the Telegram bot, deploying to Fly.io/Netlify, hatching blockchain buddies, running OODA loops, configuring voice mode (ElevenLabs + Grok), minting Metaplex agents, managing the vault, running the worker swarm, or contributing to the project. Covers all 31 MCP tools, 18 buddy species, 9 spinners, 60+ Telegram commands, 95 skills, and the full repo structure.

solana-clawd — The Agentic Engine Solana Deserves

Repo: github.com/x402agent/solana-clawd Live: solanaclawd.com npm: solana-clawd (v1.6.0) Token: $CLAWD — 8cHzQHUS2s2h8TzCmfqPKYiM4dSt4roa3n7MyRLApump

When to Use

  • Setting up solana-clawd from scratch (clone, install, configure)
  • Connecting MCP tools to Claude Desktop, Cursor, VS Code
  • Starting the Telegram trading bot
  • Deploying the web app or MCP server
  • Hatching blockchain buddies or running spinners
  • Running OODA trading loops
  • Running Phoenix perpetuals through the official Solana CLAWD perps agent and bundled Vulcan skills
  • Configuring voice mode (ElevenLabs + Grok)
  • Minting agents on Metaplex
  • Managing encrypted vault secrets
  • Contributing new tools, buddies, or skills

One-Shot Install

git clone https://github.com/x402agent/solana-clawd
cd solana-clawd
npm run setup

npm run setup is the canonical bootstrap. It installs dependencies, builds the root runtime, builds MCP/, builds packages/agentwallet/, builds the main web app, builds the Clawd Vault web app, builds the wiki app, syncs the skills catalog, and creates .env from .env.example if needed.

Minimum keys needed for live Solana data: HELIUS_API_KEY (free at helius.dev).

Quick Demo (no keys needed)

npm run demo
npm run birth
npm run spinners

Install As A Standalone Skill

Install only the master solana-clawd skill:

npx skills add x402agent/solana-clawd --path skill/solana-clawd

Install the full Solana-clawd skill pack from this repo:

npx skills add x402agent/solana-clawd

MCP Setup

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "solana-clawd": {
      "command": "node",
      "args": ["/path/to/solana-clawd/MCP/dist/index.js"],
      "env": { "HELIUS_API_KEY": "your-key" }
    }
  }
}

Cursor / VS Code

{
  "solana-clawd": {
    "command": "node",
    "args": ["MCP/dist/index.js"],
    "cwd": "/path/to/solana-clawd"
  }
}

Public HTTP (no install)

{
  "solana-clawd": { "type": "http", "url": "https://solana-clawd.fly.dev/mcp" }
}

MCP Tools (31)

CategoryToolsDescription
Market Data (8)solana_price, solana_trending, solana_token_info, solana_wallet_pnl, solana_search, solana_top_traders, solana_wallet_tokens, sol_priceLive prices, trending, security scores, smart money, PnL
Helius Onchain (8)helius_account_info, helius_balance, helius_transactions, helius_priority_fee, helius_das_asset, helius_webhook_create, helius_webhook_list, helius_listener_setupRPC, DAS metadata, enhanced txs, webhooks, WebSocket
Agent Fleet (3)agent_spawn, agent_list, agent_stopSpawn research/OODA/scanner/dream agents
Memory (2)memory_recall, memory_writeKNOWN/LEARNED/INFERRED 3-tier memory
Metaplex (6)metaplex_mint_agent, metaplex_register_identity, metaplex_read_agent, metaplex_delegate_execution, metaplex_verify_mint, metaplex_agent_walletMint AI agents as MPL Core assets
Pump.fun (8)pump_token_scan, pump_buy_quote, pump_sell_quote, pump_graduation, pump_market_cap, pump_top_tokens, pump_new_tokens, pump_cashback_infoBonding curve scanner, quotes, graduation
Skills (2)skill_list, skill_readBrowse and load 95 SKILL.md files

Phoenix Perps: Vulcan / Rise SDK

The repo includes the full Vulcan skill pack under vulcan-cli-master/skills and exposes forwarding skills under skills/vulcan-*.

Use this loading order for Phoenix perps work:

  1. skills/vulcan/SKILL.md
  2. vulcan-cli-master/CONTEXT.md
  3. vulcan-cli-master/skills/INDEX.md
  4. Focused skills such as vulcan-market-intel, vulcan-risk-management, vulcan-trade-execution, vulcan-twap-execution, vulcan-grid-trading, or vulcan-ta-strategy

Official Python perps agent:

cd solana-python-agent
python perps_agent.py health
python perps_agent.py market SOL
python perps_agent.py paper-order buy SOL --notional-usdc 100
python perps_agent.py preflight

Safety contract:

  • Paper mode is default.
  • Live orders require explicit --yes.
  • Live strategies require preflight and explicit mode selection.
  • The agent never reads or prints private keys, wallet passwords, or MCP config secrets.

Telegram Bot

# Set minimum env vars
export TELEGRAM_BOT_TOKEN=<from-botfather>
export HELIUS_API_KEY=<from-helius.dev>
export TELEGRAM_ALLOWED_CHATS=<your-chat-id>  # lock to your user

# Start
npx tsx src/telegram/index.ts

Key Commands

CategoryCommands
Market/sol, /price <mint>, /trending, /token <mint>, /wallet <addr>, /market, /latest, /graduated
Research/research <mint>, /deepresearch <mint>, /ooda, /chart <mint>, /trades <mint>, /toptraders <mint>, /holders <mint>
Pump.fun/scan, /snipe, /stop, /signal, /grad <mint>, /mcap <mint>, /cashback <mint>
Helius/balance, /tokens, /txs, /slot, /assets
AI/grok <q>, /xsearch <q>, /wsearch <q>, /imagine <prompt>, /video <prompt>, /vision <url>
Twitter/tweet <text>, /reply <id> <text>, /like <id>, /rt <id>, /tsearch <q>, /autotweet on
Vault/vault, /vault store <label> <secret>, /vault get <id>, /vault lock
System/status, /agentstate, /skills, /help, /watch <mint>

Voice Mode

The web app at solanaclawd.com/voice supports two providers:

ElevenLabsGrok (xAI)
AgentConversational AI (signed URL)Realtime API (ephemeral tokens, server VAD)
TTSRoger, Sarah, River, WillRex, Eve, Ara, Sal, Leo + speech tags
Tools--Web search during conversation
# Required env vars (web/.env)
ELEVEN_LABS_API_KEY=<elevenlabs-key>
ELEVENLABS_AGENT_ID=<agent-id>
XAI_API_KEY=<xai-key>

Blockchain Buddies

18 species across 5 categories. 8 stats per buddy. 5 rarity tiers (common 60% to legendary 1%). Multi-frame ASCII sprites with 12 hat options.

import { createBlockchainBuddy, formatBuddyCard } from 'solana-clawd/buddy'
const buddy = createBlockchainBuddy()
console.log(formatBuddyCard(buddy))

Species: SolDog, BONK Dog, dogwifhat, Jupiter Agg, Raydium LP, Whale, Bull, Bear, MEV Shark, Octopus, DeGod, y00t, Okay Bear, Pepe, Pump.fun, Sniper Bot, Validator, RPC Node.

OODA Trading Loop

npx tsx examples/ooda-loop.ts
OBSERVE  -> sol_price, trending, helius_priority_fee, memory KNOWN
ORIENT   -> score candidates (trend + momentum + liquidity + participation - risk)
DECIDE   -> confidence >= 60? -> size band
ACT      -> trade_execute gated at `ask` permission
LEARN    -> write INFERRED -> Dream agent promotes to LEARNED

9 built-in agents: Explore, Scanner, PumpScanner, SniperBot, OODA, Dream, Analyst, Monitor, MetaplexAgent.

Metaplex Agent Minting

import { createAgentUmiFromEnv, mintClawdAgentFromTemplate } from 'solana-clawd/metaplex'
const umi = createAgentUmiFromEnv('solana-devnet')
const result = await mintClawdAgentFromTemplate(umi, 'trader', {
  uri: 'https://arweave.net/metadata.json',
  network: 'solana-devnet',
})

Templates: explorer, scanner, trader, analyst, monitor, custom. Networks: solana-mainnet, solana-devnet, localnet, eclipse, sonic, fogo.

Vault (AES-256-GCM)

import { SolanaVault } from 'solana-clawd/vault'
const vault = await SolanaVault.create('passphrase')
await vault.store('api_key', 'sk-live-...', 'Helius key')
vault.lock()  // zero-fills key from memory

Auto-locks after 15 min. Scrypt-derived master key. Sentinel validation.

Deploy

Fly.io (MCP server)

cd MCP && fly launch --config fly.toml
fly secrets set HELIUS_API_KEY=key MCP_API_KEY=optional-bearer

Netlify (web app)

netlify deploy --prod
# Set env: ELEVEN_LABS_API_KEY, ELEVENLABS_AGENT_ID, XAI_API_KEY

Worker Swarm (iii SDK)

cd tailclawd/quickstart
iii -c iii-config.yaml      # start engine
docker compose up --build   # start 4 workers (TS + Rust + Python)

Repo Structure

MCP/               31 MCP tools, STDIO + HTTP + SSE transport
src/               58 subsystems, 400+ files (engine, agents, telegram, buddy, pump, voice...)
web/               Next.js frontend (chat, buddies, voice) — solanaclawd.com
gateway/           Express API + Telegram bot + Birdeye WebSocket
packages/          agentwallet (encrypted Solana+EVM vault SDK)
beepboop/          macOS menu bar companion (SwiftUI, Claude vision, voice)
llm-wiki-tang/     Research knowledge base (Next.js + FastAPI + Supabase)
tailclawd/         TailClawd web UI + iii SDK worker swarm
elevenlabs-mcp-main/  ElevenLabs MCP server
formal_verification/  Lean 4 risk engine spec
skills/            95 SKILL.md knowledge documents
examples/          4 runnable demos (buddies, wallet listener, OODA, x402)
docs/              architecture.md (48KB), migration guide, risk engine spec

Environment Variables

See .env.example (55 lines). Minimum:

HELIUS_API_KEY=              # Free at helius.dev (1M credits/month)

Full list by category:

CategoryKeys
CoreHELIUS_API_KEY, HELIUS_RPC_URL, SOLANA_TRACKER_API_KEY, BIRDEYE_API_KEY
LLMANTHROPIC_API_KEY, OPENROUTER_API_KEY, XAI_API_KEY
VoiceELEVEN_LABS_API_KEY, ELEVENLABS_AGENT_ID
TelegramTELEGRAM_BOT_TOKEN, TELEGRAM_CHAT_ID, TELEGRAM_ALLOWED_CHATS, TELEGRAM_ADMIN_IDS
WalletSOLANA_PUBLIC_KEY, SOLANA_PRIVATE_KEY, VAULT_PASSPHRASE
SocialCONSUMER_KEY, SECRET_KEY, ACCESS_TOKEN, ACCESS_TOKEN_SECRET, BEARER_TOKEN
DeployMCP_API_KEY, CONVEX_URL, CONVEX_DEPLOY_KEY

Pitfalls

  • No dotenv auto-load — Telegram bot needs source .env or set -a && source .env && set +a before running
  • Helius free tier — 1M credits/month, enough for dev but not high-frequency polling
  • Pump sniper — Requires SOLANA_PRIVATE_KEY for live execution; signal-only mode works without
  • Voice mode — Both ElevenLabs and xAI consume credits per request; monitor usage
  • Wiki sub-project — web/wiki/ has separate dependencies; excluded from main Next.js build via eslint.dirs
  • Node 20+ required — uses native WebSocket, structuredClone, other modern APIs
  • Bun compatible but tested primarily on Node; bun:bundle feature gates exist in voice code

Contributing

npm run build          # TypeScript compilation
npm run dev            # watch mode
npm run typecheck      # strict type checking
npm run mcp:build      # MCP server only
npm run demo           # test the demo
npm run birth          # test buddy hatching
npm run skills:catalog # regenerate skill manifest

High-impact areas: new MCP tools, buddy species, spinners, voice agent skills, Telegram commands. See CONTRIBUTING.md.

writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deploymentStars: 318kUpdated Claude CodeCodexVerifiedInstall
xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.Stars: 196kUpdated Claude CodeCodexVerifiedInstall
shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy.Stars: 118kUpdated Claude CodeCodexVerifiedInstall
childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release.Stars: 117kUpdated Claude CodeCodexVerifiedInstall