Index of every page on this site: llms.txt
All skills
base avatar

/wallet-mcp

@ef8bcd5 official
by basebase/skills120 stars
130
Run it yourself

Wallet MCP — gives your AI assistant access to a Coinbase Wallet Account via the Wallet MCP server (wallet-mcp.coinbase.com). Wallet, portfolio, sending, swapping, signing, x402 payments, batched contract calls, and transaction history across Base, Ethereum, Arbitrum, Optimism, Polygon, BNB Chain, Avalanche, and Robinhood Chain.

Use this Skill: https://skilld.dev/gh/base/skills/wallet-mcp

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencesbatch-calls.md

≈500 tokens on demand. Your agent reads this file only when SKILL.md points to it.

Batched Contract Calls (EIP-5792)

Wallet MCP exposes a batched-calls tool (typically send_calls) that submits multiple contract calls for a single user approval. Use it for arbitrary contract interactions, multi-step transactions, or any flow that combines an approval with a follow-up action.

Batching is preferred whenever a flow involves a token approval followed by a protocol action (approve + deposit, approve + supply, approve + swap, etc.). Also batch whenever a plugin or protocol endpoint returns multiple transactions in a single response. Don't split these into sequential single-send calls when one batched approval can execute them atomically.

When to use

  • Protocol interactions not covered by send or swap (DeFi, NFT mints, approvals, governance actions).
  • Combining multiple operations into one user approval.
  • Executing a transaction array returned by a plugin's prepare-style endpoint — pass the array straight through.

Shape

The MCP advertises the exact parameter names and types — defer to its tool description. The general shape is:

  • A chain string (base by default, or ethereum, arbitrum, optimism, polygon, bsc, avalanche, robinhood-chain, or the base-sepolia testnet).
  • A calls array of { to, value, data } objects:
    • to — target address, 0x-prefixed (required)
    • value — hex ETH in wei (e.g. 0x0), optional
    • data — calldata hex, optional

Approval flow

Same as any write tool: the response returns an approval URL and request ID. See approval-mode.md.

Generic orchestration

plugin or protocol API → { transactions: [...] } (or equivalent)
   ↓ map each transaction to a { to, value, data } call
batched-calls tool (chain, calls) → approval URL + request ID
   ↓ user approves
status-poll tool (request ID) → confirmed

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

1 alert14d3 checks · Risk CRITICAL
  • Gen Agent Trust Hub14d

    The skill integrates with various crypto protocols but contains severe security risks, including hardcoded API credentials and instructions to execute third-party code packages. Most importantly, it utilizes a service domain (flaunch.gg) that is blacklisted as malicious. It also instructions the AI to dynamically load additional instructions from remote servers.

  • Socket14d

    3 alerts: gptSecurity, gptAnomaly

  • Snyk14d

    Risk: MEDIUM · 2 issues

Signed by skilld at ef8bcd5. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub 2 weeks ago.

Activeupdated 2 weeks ago
title
Wallet MCP Skill
version
0.1.0

README badge

README badge for base/skills/wallet-mcp