Source profileQuality 91/100Review permissions

base/skills/skills/base-mcp/SKILL.md

base-mcp

Base MCP — gives your AI assistant access to a Base Account via the Base MCP server (mcp.base.org). Wallet, portfolio, sending, swapping, signing, x402 payments, batched contract calls, and transaction history across supported chains.

Source repository stars
114
Declared platforms
0
Static risk flags
2
Last source update
2026-08-24
Source checked
2026-08-25

Decision brief

What it does: where it fits

[!IMPORTANT] Run onboarding at the start of every conversation that touches Base MCP Including conversations that jump straight to a plugin topic. Onboarding is short — see below.

Best for

    Not for

    • Tasks that require unconfirmed production actions or broad system permissions.
    • Environments where the pinned source and install steps cannot be inspected.

    Compatibility matrix

    Platform support, with evidence labels

    PlatformStatusEvidenceWhat to check
    CodexNot declaredNo explicit evidencePortability before use
    Claude CodeNot declaredNo explicit evidencePortability before use
    CursorNot declaredNo explicit evidencePortability before use
    Gemini CLINot declaredNo explicit evidencePortability before use
    Open the compatibility checker

    Installation

    Inspect first. Install second.

    The source command is displayed only when detected. A safe inspection prompt is always available so your agent can explain every action before execution.

    Source-detected install commandSource
    npx skills add https://github.com/base/skills --skill "skills/base-mcp"
    Safe inspection promptEditorial

    Inspect the Agent Skill "base-mcp" from https://github.com/base/skills/blob/070ebc9877d4f06cd3588c6920d785a07dbb0220/skills/base-mcp/SKILL.md at commit 070ebc9877d4f06cd3588c6920d785a07dbb0220. List every install step, command, network request, credential, file read/write, external action, and rollback step. Explain whether it fits my task. Do not install or execute anything until I approve.

    Workflow

    What the source asks the agent to do

    1. 01

      Detection

      The Base MCP exposes its tools to the harness when connected. If no Base MCP tool is callable, the MCP server is not installed: direct the user to https://docs.base.org/ai-agents/quickstart (or load references/install.md for app-specific steps) and stop.

      The Base MCP exposes its tools to the harness when connected. If no Base MCP tool is callable, the MCP server is not installed: direct the user to https://docs.base.org/ai-agents/quickstart (or load references/install.m…If Base MCP tools are available, load references/tone.md — its rules apply for the entire conversation — then continue to Onboarding. If — and only if — sibling files aren't readable (e.g. you were handed just this SKIL…
    2. 02

      Onboarding

      Keep it short. Do this once per session, before doing real work:

      Briefly mention what's available — one or two sentences. The user has a Base Account wallet and can do things like check balances, send and swap tokens, sign messages, make x402 payments, batch contract calls, and (if i…Show this disclaimer verbatim before proceeding:Wallet address and balance are optional — only fetch and display them when the user asks, or when a pending operation actually needs the address (e.g., a write call, a position lookup). Don't volunteer a wallet dump up…
    3. 03

      Tools

      The Base MCP advertises its own tool catalog to the harness. Read the tool descriptions exposed by the MCP — they are the source of truth and may change over time. Do not assume a fixed list; do not preload a tool catalog from this skill.

      Default — local. Read each references/… or plugins/… link from the same directory as this SKILL.md (filesystem read, e.g. Read/cat). This is the canonical source — always try this first.Fallback — web (only when local read fails). If — and only if — the local file is genuinely unreadable (e.g. you were handed SKILL.md as a pasted body with no skill directory on disk, or the filesystem read errors), the…Lazy. Only load a reference or plugin when the conversation actually needs it. Don't preload the catalog.
    4. 04

      Loading referenced files

      Default — local. Read each references/… or plugins/… link from the same directory as this SKILL.md (filesystem read, e.g. Read/cat). This is the canonical source — always try this first.

      Default — local. Read each references/… or plugins/… link from the same directory as this SKILL.md (filesystem read, e.g. Read/cat). This is the canonical source — always try this first.Fallback — web (only when local read fails). If — and only if — the local file is genuinely unreadable (e.g. you were handed SKILL.md as a pasted body with no skill directory on disk, or the filesystem read errors), the…Lazy. Only load a reference or plugin when the conversation actually needs it. Don't preload the catalog.
    5. 05

      Plugins

      Plugins extend Base MCP with third-party functionality (lending, swaps, perps, etc.). Each plugin is a separate third-party service that Base does not operate, endorse, or audit. The available set may change and users might drop additional instructions in the chat or custom plug…

      Plugins extend Base MCP with third-party functionality (lending, swaps, perps, etc.). Each plugin is a separate third-party service that Base does not operate, endorse, or audit. The available set may change and users m…Open a plugin only when the user names that specific third-party platform (e.g. "swap on Uniswap", "lend on Moonwell", "buy a gift card on Bitrefill") or otherwise unambiguously points to it. Never infer a platform from…If the user asks for a capability without naming a platform — "swap some USDC", "earn yield", "launch a token", "buy a gift card" — do not pick a plugin for them. Instead, list the plugins from the table below that cove…

    Permission review

    Static risk signals and limitations

    Reads files

    low · line 13

    The documentation asks the agent to read local files, directories, or repositories.

    If Base MCP tools are available, load [references/tone.md](references/tone.md) — its rules apply for the entire conversation — then continue to Onboarding. If — and only if — sibling files aren't readable (e.g. you were handed just this `SK

    Reads files

    low · line 44

    The documentation asks the agent to read local files, directories, or repositories.

    **Fallback — web (only when local read fails).** If — and only if — the local file is genuinely unreadable (e.g. you were handed `SKILL.md` as a pasted body with no skill directory on disk, or the filesystem read errors), then fetch the sam

    Runs scripts

    medium · line 91

    The documentation asks the agent to run terminal commands or scripts.

    npx skills add base/skills --skill base-mcp

    Evidence record

    Why each signal appears

    EvidenceSourceComputedTestedEditorial
    SignalValueEvidence typeMeaning
    Quality score91/100ComputedDocumentation, specificity, maintenance, and trust rules
    Repository stars114SourceRepository attention, not individual Skill quality
    Compatibility0 platformsSourceDeclared in the catalog source record
    Usage guideautomated source guideEditorialGenerated or reviewed according to the visible evidence level

    Pinned source

    Provenance and original SKILL.md

    Repository
    base/skills
    Skill path
    skills/base-mcp/SKILL.md
    Commit
    070ebc9877d4f06cd3588c6920d785a07dbb0220
    License
    MIT
    Collected
    2026-08-25
    Default branch
    master
    View the original SKILL.md

    Base MCP

    [!IMPORTANT]

    Run onboarding at the start of every conversation that touches Base MCP

    Including conversations that jump straight to a plugin topic. Onboarding is short — see below.

    Detection

    The Base MCP exposes its tools to the harness when connected. If no Base MCP tool is callable, the MCP server is not installed: direct the user to https://docs.base.org/ai-agents/quickstart (or load references/install.md for app-specific steps) and stop.

    If Base MCP tools are available, load references/tone.md — its rules apply for the entire conversation — then continue to Onboarding. If — and only if — sibling files aren't readable (e.g. you were handed just this SKILL.md body and have no local filesystem access to the skill directory), fetch the same relative path from https://docs.base.org/ai-agents/skills/references/tone.md using the web_request tool from Base MCP. The same fallback applies to every other reference and plugin link in this file (see "Loading referenced files" below).

    Onboarding

    Keep it short. Do this once per session, before doing real work:

    1. Briefly mention what's available — one or two sentences. The user has a Base Account wallet and can do things like check balances, send and swap tokens, sign messages, make x402 payments, batch contract calls, and (if installed) use third-party plugins for DeFi, swaps, and other onchain actions. Do not enumerate every tool — the agent discovers tools and plugins directly from the MCP.

    2. Show this disclaimer verbatim before proceeding:

      By using the Base MCP, you agree to the Base Account and Base App Terms of Service. Base MCP provides access to plugins that are built by third parties, not Base. Base doesn't operate, endorse, or audit them, and isn't responsible for the protocols you interact with. Transactions are irreversible — always review before approving.

    3. Wallet address and balance are optional — only fetch and display them when the user asks, or when a pending operation actually needs the address (e.g., a write call, a position lookup). Don't volunteer a wallet dump up front.

    Tools

    The Base MCP advertises its own tool catalog to the harness. Read the tool descriptions exposed by the MCP — they are the source of truth and may change over time. Do not assume a fixed list; do not preload a tool catalog from this skill.

    Two patterns deserve their own references because they span multiple tools:

    TopicReference
    Approval flow (for any write tool that returns an approval URL)references/approval-mode.md
    Batched contract calls (EIP-5792)references/batch-calls.md
    Custom / non-native plugins and the web_request allowlistreferences/custom-plugins.md
    Platform install stepsreferences/install.md
    Tone and language rulesreferences/tone.md

    Loading referenced files

    • Default — local. Read each references/… or plugins/… link from the same directory as this SKILL.md (filesystem read, e.g. Read/cat). This is the canonical source — always try this first.
    • Fallback — web (only when local read fails). If — and only if — the local file is genuinely unreadable (e.g. you were handed SKILL.md as a pasted body with no skill directory on disk, or the filesystem read errors), then fetch the same relative path from https://docs.base.org/ai-agents/skills/<path> (replace <path> with the relative link, e.g. references/tone.md, plugins/morpho.md). Use the web_request tool from Base MCP for this fetch — that is the required mechanism on every surface, including ones with a built-in web/fetch tool. Do not skip this fallback and do not try to operate from memory or guesses — if the local read fails, you must web_request the file before acting on its contents.
    • Lazy. Only load a reference or plugin when the conversation actually needs it. Don't preload the catalog.

    Plugins

    Plugins extend Base MCP with third-party functionality (lending, swaps, perps, etc.). Each plugin is a separate third-party service that Base does not operate, endorse, or audit. The available set may change and users might drop additional instructions in the chat or custom plugins that would allow you to use other protocols with the MCP.

    Routing rule — name-gated, never auto-routed

    Open a plugin only when the user names that specific third-party platform (e.g. "swap on Uniswap", "lend on Moonwell", "buy a gift card on Bitrefill") or otherwise unambiguously points to it. Never infer a platform from a bare capability — choosing one third-party service over another is the user's decision, not yours.

    If the user asks for a capability without naming a platform — "swap some USDC", "earn yield", "launch a token", "buy a gift card" — do not pick a plugin for them. Instead, list the plugins from the table below that cover that capability (and mention any native Base MCP tool that also does, e.g. the built-in swap), then let the user choose. Only after the user names a platform do you open its plugin and act.

    Plugins currently maintained alongside this skill (the native plugins). Use the table two ways: (a) to route a request that already names a platform to the right plugin, and (b) to assemble the option list when the user asks for a capability without naming one. The "Open it when…" column intentionally anchors each trigger to the platform's own name — that name is the trigger.

    PluginOpen it when the user names this platform and wants to...Common actions it coversNotice before acting
    AerodromeSwap, provide liquidity, stake, or claim rewards on Aerodrome.Pool discovery, swaps, LP add/remove, staking, unstaking, reward claims.Base only. Requires a shell-capable harness for the Sugar CLI; stop on chat-only surfaces.
    AvantisTrade leveraged perpetual futures or inspect Avantis trading activity.Open/close positions, cancel orders, adjust margin, set TP/SL, view positions, history, and PnL.Base only. Leverage can liquidate the position; chat-only surfaces can read data but use the Avantis UI for trade actions.
    BalancerSwap tokens or provide liquidity on Balancer.Read pools/quotes, build swap calldata, add/remove liquidity, submit via send_calls.CLI-only — requires shell access; stop on chat-only surfaces. Multi-chain (Base, Ethereum, Arbitrum, Optimism, Avalanche); confirm slippage and low-liquidity risk.
    BankrDiscover fresh token launches or buy a newly launched token on Bankr.Read launch feeds, inspect token metadata, buy a selected token.Base only. New tokens can be illiquid or unsafe; do not auto-buy, and make price/liquidity risk explicit.
    BitrefillBuy gift cards, mobile top-ups, or travel eSIMs on Bitrefill, paid with USDC.Search 1,500+ brands, check out, deliver gift-card codes and eSIM details in chat.Base only. Requires SIWE login; purchases are irreversible and handle personal data — confirm order details before paying.
    BrickkenManage ERC-8004 agent identity, reputation, or agent-token operations on Brickken.Identity and reputation reads/writes, agent-token operations via x402 approval.Base and Base Sepolia. Writes are irreversible and may involve personal data; review before approving.
    ClawnchDiscover recent/top token launches or launch a token on Clawnch.Browse launch feeds, inspect tokens, non-custodial launch, swap via send_calls.Base only. New tokens can be illiquid or unsafe and launches are irreversible; make risk explicit.
    FlaunchLaunch a token on Flaunch or discover/swap deployed Flaunch tokens.Prepare launches via mcp.flaunch.gg, discover tokens, swap, submit via send_calls.Base only. New tokens can be illiquid; launches and swaps are irreversible — confirm details first.
    GMGNGet token swap quotes, gas prices, or trending-token market intelligence on Base via GMGN.Swap quotes (unsigned calldata via send_calls), gas-price tiers, trending tokens.Base only. CLI-only — requires shell to generate auth params; API key auth. Confirm slippage and watch low-liquidity tokens.
    HydrexSwap tokens or manage concentrated-liquidity positions on Hydrex.Quotes, swaps, concentrated-liquidity add/remove, submit via send_calls.Base only. Quotes can move; confirm slippage before writes.
    KyberSwapSwap tokens on KyberSwap (aggregated best-rate routing across chains).Best-rate routing through 50+ liquidity sources, quotes, swaps via send_calls.Multi-chain (Base, Ethereum, Arbitrum, Optimism, Polygon, BSC, Avalanche); quotes can move, confirm slippage before swapping.
    MoonwellLend, borrow, repay, withdraw, or check Moonwell positions/rewards.Market/rate reads, health checks, supply, withdraw, borrow, repay, rewards lookup.Works on Base and Optimism. Borrowing can be liquidated; surface health factor before risky actions.
    MorphoEarn yield, use Morpho vaults, or borrow/lend in Morpho markets.Compare vaults, deposit or withdraw, supply collateral, borrow, repay, check positions.Base only. Borrowing can be liquidated; read health/position data before preparing writes.
    o1.exchangeSwap tokens on o1.exchange, optionally with Permit2 gasless approvals and MEV-protected routing.Quotes, standard swaps via send_calls, Permit2 swaps via private relay.Multi-chain (Base, BSC). Quotes can move and swaps are irreversible; confirm slippage and watch for low-liquidity tokens.
    OpenSeaTrade NFTs, swap tokens, or mint drops on OpenSea.NFT buy/sell/list, token swaps, drops/minting, submit via send_calls.Multi-chain (Ethereum, Base, Polygon, Arbitrum, Optimism, Avalanche). Requires an API key; trades are irreversible — confirm token, price, and slippage first.
    PrintrLaunch a cross-chain token on Printr.Prepare token-creation calldata via HTTP API, submit via send_calls.Multi-chain (Base, Arbitrum, Optimism, Polygon, BSC, Avalanche, Ethereum). New tokens can be illiquid; launches are irreversible.
    UniswapSwap tokens or manage Uniswap liquidity positions.Token swaps, approval checks, V2/V3/V4 LP create/increase/decrease/collect flows.Base only. Quotes can move; confirm slippage and token/position details before writes.
    VeniceRun private AI inference or media generation on Venice, optionally funded with x402.Text and media inference via the Venice API, optional Base x402 wallet funding.Base only. Requires SIWE login; inference handles personal data and paid calls are irreversible.
    VirtualsCreate or operate Virtuals AI agents, payment cards, or agent email.Agent management, card setup and limits, email inbox/thread actions, SIWE login.Requires the Virtuals MCP and a signed login. Handles personal data; avoid exposing tokens, OTPs, card details, or email contents unnecessarily.
    YOView YO vaults, check positions, deposit, or request redeem on YO yield vaults.ERC-4626 vault reads via chain_rpc_request, deposit, request redeem via send_calls.Multi-chain (Base, Ethereum, Arbitrum). Deposits and redeems are irreversible; confirm amounts and vault before writes.

    Load a plugin reference only once the user has named the platform it covers (per the routing rule above), following the same local-first, web-fallback rule as references (see Loading referenced files above). For a plugin's own external tools, defer to the plugin file first, then to any CLI help, API schema, or MCP tool descriptions it explicitly tells you to use.

    Native plugins vs. custom / user-supplied plugins

    Native plugin HTTP hosts may be allowlisted in the Base MCP web_request tool. Aerodrome, Balancer, and GMGN are CLI-only and require a harness with shell access. Avantis is hybrid: view-only reads (market data, positions, PnL) work on every surface via web_request, while tx-builder calls require a CLI harness — on chat-only surfaces the plugin links the user to the Avantis web UI instead (see plugins/avantis.md). Morpho is hybrid too: use Morpho CLI when shell access exists, otherwise use or install the Morpho MCP as described in plugins/morpho.md. Custom or user-supplied plugins usually aren't allowlisted — load references/custom-plugins.md for the decision tree on which HTTP path to use (harness HTTP tool vs. user-paste fallback, and the GET-only constraint on Claude/ChatGPT consumer surfaces).

    Installation

    npx skills add base/skills --skill base-mcp
    

    Frequently asked questions

    What to verify before installation and use

    What does the base-mcp source document cover?

    [!IMPORTANT] Run onboarding at the start of every conversation that touches Base MCP Including conversations that jump straight to a plugin topic. Onboarding is short — see below.

    How do I install base-mcp?

    The source record exposes this install command: npx skills add https://github.com/base/skills --skill "skills/base-mcp". Inspect the command and pinned source before running it.

    Which permission-related actions were detected?

    Static rules flagged read-files, exec-script in the source; the page lists the matching lines and excerpts.