CCM
/Skills
SkillsMCPMarketplacesDigestToolsAdvertise

This week in Claude

Every Monday: Claude Code, Agent SDK, MCP, and the Anthropic platform moves worth your time.

Skills by Category
Frontend DevelopmentBackend & APIsTesting & QASecurityDevOps & CI/CDGit & Pull RequestsDocumentationCode Review & QualityAI & Agent BuildingSkill Development
MCP Servers by Category
Sales & MarketingWeb & Browser AutomationDatabasesAI & LLM ToolsCloud & InfrastructureCommunication & MessagingDeveloper ToolsDesign & CreativeDocuments & KnowledgeSearch & Web Crawling
Marketplaces by Category
AI Agents & OrchestrationLLM IntegrationDevelopment ToolsFrontend & UIBackend & APIsDatabasesTesting & Code QualityDevOps & CloudSecurity & ComplianceGit & Version Control

Claude Code Marketplaces

Discover Claude Code plugins, extensions, and tools. Automatically updated directory of Anthropic Claude AI marketplaces with development tools, productivity plugins, and integrations.

Resources

  • Browse Skills
  • Browse MCP Servers
  • Browse Marketplaces
  • Plugins Reference

Community

  • About
  • Tools
  • Feedback
  • Privacy Policy
  • Advertise

Built for the Claude Code community with Claude Code by mertbuilds.com

Independent project, not affiliated with Anthropic

Wallet

starchild-ai-agent/official-skills
8.4k installs18 stars
Summary

A proper multi-chain wallet implementation that handles the messy reality of blockchain operations. Covers EVM chains via DeBank and Solana via Birdeye, with gas sponsorship that falls back to user-paid when needed. The policy system is notably well-documented since Privy's rule precedence (DENY always wins) trips up most implementations. Includes all the signing primitives you'd expect: EIP-191, EIP-712, transaction signing, plus broadcast capabilities. The three policy modes (allow-all, deny-all, whitelist) cover the main security patterns without the usual footguns. Solid choice if you need wallet functionality that actually works in production rather than just demos.

Install to Claude Code

npx -y skills add starchild-ai-agent/official-skills --skill wallet --agent claude-code

Installs into .claude/skills of the current project.

CodeRabbit
CodeRabbit
AI writes the code. CodeRabbit catches the slop.
Try For Free →
Email for Agents: Free tier availableEmail for Agents: Free tier available
Email for Agents: Free tier available
Give your AI agent a complete email layer—sending, inbound inboxes, and sandbox testing.
Get 4K emails/month free →
Make your agent a DeFi expert
Make your agent a DeFi expert
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
AppSignal
AppSignal
Monitor with ease. Code with confidence.
Start Free Trial →
Capacitor - Shared memory for your team’s coding agents.
Capacitor - Shared memory for your team’s coding agents.
Capacitor makes coding agent sessions - Searchable, Shareable, Vendor-neutral & Scored.
Try For Free →
Give your AI the whole web as clean markdownGive your AI the whole web as clean markdown
Give your AI the whole web as clean markdown
Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →
belt - the only tool your agent needs
belt - the only tool your agent needs
belt cli automatically finds the best tools and skills for your agent. image, video, music, tts...
one prompt install →
inference shell
inference shell
create and run specialised agents in minutes
build now →
CodeRabbit
CodeRabbit
AI writes the code. CodeRabbit catches the slop.
Try For Free →
Email for Agents: Free tier availableEmail for Agents: Free tier available
Email for Agents: Free tier available
Give your AI agent a complete email layer—sending, inbound inboxes, and sandbox testing.
Get 4K emails/month free →
Make your agent a DeFi expert
Make your agent a DeFi expert
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
AppSignal
AppSignal
Monitor with ease. Code with confidence.
Start Free Trial →
Capacitor - Shared memory for your team’s coding agents.
Capacitor - Shared memory for your team’s coding agents.
Capacitor makes coding agent sessions - Searchable, Shareable, Vendor-neutral & Scored.
Try For Free →
Give your AI the whole web as clean markdownGive your AI the whole web as clean markdown
Give your AI the whole web as clean markdown
Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →
belt - the only tool your agent needs
belt - the only tool your agent needs
belt cli automatically finds the best tools and skills for your agent. image, video, music, tts...
one prompt install →
inference shell
inference shell
create and run specialised agents in minutes
build now →
Files
  • logo.png
SKILL.mdView on GitHub

💰 Wallet Skill

Multi-chain wallet for EVM (DeBank-supported chains) + Solana. Balances, transfers, signing, and policy management. Script skill — call the functions below via bash; no wallet tools are registered.

How to call

All read/transfer/sign operations are Python functions in core.skill_tools.wallet. Run them from bash and read the JSON result:

python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.wallet_balance(chain='base')))"

The one operation that is NOT a script function is proposing a wallet policy — it needs to render a confirmation card in the UI, so it goes through the native frontend_action tool (see Policy Management below).

Functions (from core.skill_tools import wallet)

FunctionDescription
wallet_info()Get all AGENT wallet addresses
get_user_wallets()The USER'S OWN wallets (login + secondary) — read-only, from env
wallet_balance(chain, address="", asset="")EVM balance on a chain (DeBank). chain required
wallet_sol_balance(address="", asset="")Solana balance (Birdeye)
wallet_get_all_balances(evm_address="", sol_address="")All chains at once
wallet_transfer(to, amount, chain_id=1, data="", **kw)Broadcast EVM tx (gas sponsored by default)
wallet_sign_transaction(to, amount, chain_id=1, data="", **kw)Sign EVM tx (no broadcast)
wallet_sign(message)EIP-191 message signing
wallet_sign_typed_data(domain, types, primaryType, message)EIP-712 typed data signing
wallet_transactions(chain="ethereum", asset="", limit=20)EVM tx history
wallet_sol_transfer(transaction, caip2=...)Broadcast Solana tx (base64)
wallet_sol_sign_transaction(transaction)Sign Solana tx (no broadcast)
wallet_sol_sign(message)Solana message signing
wallet_sol_transactions(chain="solana", asset="sol", limit=20)Solana tx history
wallet_get_policy(chain_type="ethereum")Check policy status
validate_and_clean_rules(rules, chain_type)Pre-validate policy rules before proposing

The User's Own Wallets (login / secondary)

The agent wallet is NOT the user's wallet. The platform injects the user's own wallet identities as env vars (synced from the control plane at container start and on user wallet actions):

  • USER_LOGIN_WALLET_ADDRESS / USER_LOGIN_WALLET_TYPE — the wallet the user logs in with (or bound as primary).
  • USER_SECONDARY_WALLET_ADDRESS / USER_SECONDARY_WALLET_TYPE — the user's other linked wallet (e.g. Solana when login is EVM).

When asked "what's my wallet" / "my login wallet" / "check MY balance", read these — do NOT answer with the agent wallet or say you don't know:

python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.get_user_wallets()))"

Empty/missing values mean the user has never bound a wallet in that slot (e.g. social login) — say so and point them to wallet binding in the web app.

Rules:

  • Read-only. The agent holds no keys for these wallets. To check the user's balances, pass the address into wallet_balance(chain, address=...) / wallet_sol_balance(address=...).
  • Transactions from the user's wallet never go through script functions — use the native frontend_action(action_type="user_wallet_tx", ...) flow, where the user signs in the UI and expected_from is enforced server-side.

Key Facts

  • Amounts are in wei for EVM (wallet_transfer / wallet_sign_transaction). 0.01 ETH = 10000000000000000. For ERC-20 token sends, amount is 0 (native) and the transfer is encoded in data calldata.
  • Gas is sponsored by default on EVM chains — user doesn't need native tokens for gas. Falls back to user-paid if unavailable. Pass sponsor=False to pay gas from wallet balance.
  • Policy default: OFF (allow-all). Only when policy is enabled do transactions need UI confirmation.
  • Supported EVM chains: All DeBank-supported chains. Common names auto-mapped (e.g. avalanche → avax, bsc → bsc, zksync → era). Fallback aliases include ethereum/base/arbitrum/optimism/polygon/linea/bsc/avalanche/fantom/gnosis/zksync/scroll/blast/mantle/celo/aurora plus monad/world/unichain/abstract/sonic/berachain.
  • Balance sources: DeBank (EVM), Birdeye (Solana), wallet-service (fallback). DeBank/Birdeye keys are auto-injected by sc-proxy.

Workflows

Check balances

python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.wallet_balance(chain='base')))"
python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.wallet_get_all_balances()))"

Send a transaction (EVM)

Always verify balance before, and the result/history after.

# 1. check
python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.wallet_balance(chain='base')))"
# 2. transfer (amount in wei)
python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.wallet_transfer(to='0x...', amount='10000000000000000', chain_id=8453)))"
# 3. verify
python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.wallet_transactions(chain='base')))"

Sign EIP-712 typed data

python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.wallet_sign_typed_data(domain={...}, types={...}, primaryType='Permit', message={...})))"

Policy Management

Checking policy is a script function; proposing a policy uses the native frontend_action tool (it renders a signature card in the UI — a script cannot).

  1. Check current policy:
    python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.wallet_get_policy(chain_type='ethereum')))"
    
  2. (Optional) pre-validate rules:
    python3 -c "from core.skill_tools import wallet; import json; print(json.dumps(wallet.validate_and_clean_rules([...], 'ethereum')))"
    
  3. Propose — call the frontend_action tool (not a script):
    frontend_action(action_type="update_wallet_policy", chain_type="ethereum", rules=[...])
    
    The user confirms + signs in the UI. Call once per chain (EVM + Solana = two calls).

Standard Wildcard Policy (when needed)

rules = [
  {"name": "Deny key export", "method": "exportPrivateKey", "conditions": [], "action": "DENY"},
  {"name": "Allow all", "method": "*", "conditions": [], "action": "ALLOW"},
]

Policy Modes — CRITICAL DECISION TABLE

⚠️ DENY > ALLOW in Privy. DENY * overrides ALL ALLOW rules. NEVER mix them.

ModeRulesEffect
Allow-all (default)DENY exportPrivateKey + ALLOW *Everything allowed except key export
Deny-all (lockdown)DENY exportPrivateKey + DENY *Nothing works. No ALLOW rules!
Whitelist (selective)DENY exportPrivateKey + specific ALLOW rules onlyOnly whitelisted ops work, rest implicitly denied

Mode 1: Allow-All (Standard Wildcard)

rules = [
  {"name": "Deny key export", "method": "exportPrivateKey", "conditions": [], "action": "DENY"},
  {"name": "Allow all", "method": "*", "conditions": [], "action": "ALLOW"},
]

Mode 2: Deny-All (Lockdown)

rules = [
  {"name": "Deny key export", "method": "exportPrivateKey", "conditions": [], "action": "DENY"},
  {"name": "Deny all actions", "method": "*", "conditions": [], "action": "DENY"},
]
# ⚠️ NO ALLOW rules here — DENY * would override them!

Mode 3: Whitelist (Selective Allow)

rules = [
  {"name": "Deny key export", "method": "exportPrivateKey", "conditions": [], "action": "DENY"},
  {"name": "Allow transfer to Uniswap", "method": "eth_sendTransaction", "conditions": [
    {"field_source": "ethereum_transaction", "field": "to", "operator": "eq", "value": "0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913"}
  ], "action": "ALLOW"},
]
# ⚠️ NO "DENY *" here! enabled=true already denies everything not ALLOWed.
# Adding DENY * would override the ALLOW rules above (DENY > ALLOW).

Privy Policy Rules — Key Constraints

RuleDetails
Default behaviorenabled=true → deny-all unless explicitly ALLOWed
DENY > ALLOWDENY always wins when both match
Empty conditionsOnly exportPrivateKey and * (wildcard) allow conditions: []
TX methods need conditionseth_sendTransaction, eth_signTransaction, eth_signTypedData_v4, eth_signUserOperation, signAndSendTransaction, etc. ALL require ≥1 condition
Valid field_sourcesEVM: ethereum_transaction (to/value/chain_id), ethereum_calldata (function_name), ethereum_typed_data_domain (chainId/verifyingContract), ethereum_typed_data_message, system
Valid operatorseq, gt, gte, lt, lte, in (array, max 100 values)
Dual chainCall frontend_action(action_type="update_wallet_policy", ...) TWICE for EVM + Solana

Gotchas

  • Policy proposal goes through the frontend_action tool — needs an active SSE session (won't work from a background task).
  • wallet_balance requires chain — use wallet_get_all_balances for discovery.
  • For both EVM + Solana policy, call frontend_action TWICE (one per chain_type).
Featured
CodeRabbit
CodeRabbit
AI writes the code. CodeRabbit catches the slop.
Try For Free →
Email for Agents: Free tier availableEmail for Agents: Free tier available
Email for Agents: Free tier available
Give your AI agent a complete email layer—sending, inbound inboxes, and sandbox testing.
Get 4K emails/month free →
Make your agent a DeFi expert
Make your agent a DeFi expert
Agent, run crypto. Access onchain data & trade routes via 1inch.
Install now →
AppSignal
AppSignal
Monitor with ease. Code with confidence.
Start Free Trial →
Capacitor - Shared memory for your team’s coding agents.
Capacitor - Shared memory for your team’s coding agents.
Capacitor makes coding agent sessions - Searchable, Shareable, Vendor-neutral & Scored.
Try For Free →
Give your AI the whole web as clean markdownGive your AI the whole web as clean markdown
Give your AI the whole web as clean markdown
Integrate web data into your AI product. One API to scrape website & brand data.
Get API Key Now →
belt - the only tool your agent needs
belt - the only tool your agent needs
belt cli automatically finds the best tools and skills for your agent. image, video, music, tts...
one prompt install →
inference shell
inference shell
create and run specialised agents in minutes
build now →
Categories
AI & Agent BuildingWeb3 & Blockchain
View on GitHub

Recommended

More AI & Agent Building →
agent-memory-mcp

sickn33/antigravity-awesome-skills

agent memory mcp
1.2k
43.1k
agent-memory-mcp

davila7/claude-code-templates

agent memory mcp
569
29.4k
llm-application-dev-langchain-agent

sickn33/antigravity-awesome-skills

llm application dev langchain agent
306
39.4k
llm-application-dev

moizibnyousaf/ai-agent-skills

Building applications with Large Language Models - prompt engineering, RAG patterns, and LLM integration. Use for AI-powered features, chatbots, or LLM-based automation.
1.1k
ai-prompt-engineering-safety-review

github/awesome-copilot

Comprehensive safety analysis and improvement framework for AI prompts with detailed assessment methodologies.
9.8k
36.5k
emblem-ai-prompt-examples

emblemcompany/agent-skills

emblem ai prompt examples
8.8k
12