foxd-mcp
Installed, started with no arguments, no environment and no input, and listed its tools. Run 2026-11-16.
Needs to start: nothing
Returns a result when called with no setup: 0 of 5 tools tried
Contacts while installing and starting: foxd.xyz
What it contacted
foxd.xyz
Hosts it looked up or connected to while installing and starting with no input, watched from outside its sandbox. The npm registry is left out. Says nothing about what its tools contact when called.
Starting it
npx -y foxd-mcp@0.3.1
Installing a package runs its code. Use a sandbox for anything you have not read.
Calling its tools
5 tools that take no input and look read-only were called once with nothing. 0 returned a result.
copyIntel.categories | No usable answer tool reported an error: [api_key_required] copyIntel.categories requires an API key with scope: read |
wallet.list | No usable answer tool reported an error: [api_key_required] wallet.list requires an API key with scope: wallet:read |
alerts.list | No usable answer tool reported an error: [api_key_required] alerts.list requires an API key with scope: read:account |
account.whoami | No usable answer tool reported an error: [api_key_required] account.whoami requires an API key with scope: read:account |
keys.list | No usable answer tool reported an error: [api_key_required] keys.list requires an API key with scope: keys:manage |
During the calls it contacted nothing it had not already contacted.
Only tools whose names read like looking something up, with no required input, are tried, five at most. Tools that need input, and anything that sounds like it changes or sends something, are not called.
What its tools tell an agent
In the tool descriptions, nothing that talks about the user, other tools or private files, or uses hidden characters.
1 tool with commands in capitals or very long text
- Gives the agent orders in
wallet.portfolio:ack with wallet.reclaimRent. A token with no known price is NEVER classed as dust, since unpriced does not mean worthless. `degraded: true` means a source failed and
Common and usually harmless: it is how many makers stress how a tool should be used.
Found by simple pattern rules in each tool's description (up to 3,000 characters), quoted as written. Parameter descriptions are not scanned. A clean result is not a safety review.
Tools
auth.challenge address
Step one of headless wallet sign-in. Returns a single-use nonce and the EXACT message string to sign — sign that string verbatim, byte for byte, because any modification invalidates the signature. Works for both EVM (personal_sign / EIP-191) and Solana (ed25519 signMessage). The challenge expires in 10 minutes and can be redeemed only once. Follow with auth.verifyWallet. Public — no API key required.
auth.verifyWallet message, signature, address, keyName, scopes
Step two of headless wallet sign-in: submit the signature over the challenge message. On success it creates the account if this wallet has never signed in before (provisioning one EVM and one Solana custodial wallet), or returns the existing account if it has, and issues an API key. This is the ONLY signup path that needs no human and no browser — an autonomous agent holding a keypair can go from nothing to a funded-and-tradeable account with two calls. The API key is shown once and cannot be recovered. Public — no API key required.
auth.startDeviceLogin scopes, clientName
Begin sign-in with Google, X, Telegram or email from a CLI or agent that has no browser, using the OAuth 2.0 device authorization grant — the same flow as `gh auth login`. Returns a short user code and a URL: show BOTH to the person, wait while they open it and sign in with whichever provider they prefer, then call auth.pollDeviceLogin until it completes. Use auth.verifyWallet instead when no human is available. Public — no API key required.
auth.pollDeviceLogin deviceCode
Check whether the human finished approving a device sign-in. Returns status `authorization_pending` while you wait, `slow_down` if you are polling too fast (back off by the returned interval), `access_denied` if they refused, `expired_token` if the code ran out, or `complete` with the API key and provisioned wallets. Respect the interval from auth.startDeviceLogin. The key is delivered exactly once — the poll that returns it destroys the server's copy. Public — no API key required.
auth.verifyTelegram initData, keyName, scopes
Sign in using the signed `initData` payload Telegram injects into a Mini App webview. The HMAC is verified against our bot token, so this is headless and needs no redirect. A Telegram user reaching FoxD this way resolves to the SAME account as signing in with 'Login with Telegram' on the web — one identity, one set of wallets. Only useful from inside Telegram; use auth.verifyWallet or the device flow anywhere else. Public — no API key required.
82 more tools
market.chains
List every chain FoxD indexes and trades, with native symbol, wrapped-native address, VM family (evm or svm), EVM chain id and block-explorer URL. Call this first when you need to know which chain slugs are valid — every other tool takes one of these slugs. Requires an API key with scope: read.
market.platforms chain
List every platform FoxD indexes — AMMs (Uniswap V2/V3/V4, PancakeSwap, Aerodrome, Camelot, HyperSwap, Orca, Meteora) and launchpads (pump.fun, LaunchLab/bonk.fun, four.meme, Clanker, Zora, Flaunch, alt.fun, Hypurr) — with the chain each runs on and its kind (amm_v2, amm_v3, amm_solidly, bonding, svm_amm). Filter by chain to see what is live where. Requires an API key with scope: read.
market.trending chain, sort, limit, minVolume, sparklines
Rank indexed tokens by trending score, 24h volume, recency or price change, across one chain or all six. Returns full token summaries including price, market cap, liquidity, 24h volume, buy/sell counts, holder count, safety facts and a 72h sparkline. This is the main discovery feed — use market.search when you already know what you are looking for. Requires an API key with scope: read.
market.newTokens chain, limit
The most recently discovered tokens, newest first. Broader than sniper.feed: this includes organic AMM listings as well as launchpad mints, and carries no bonding-stage bucketing. Reach for sniper.feed when you specifically want fresh launches grouped by curve stage. Requires an API key with scope: read.
market.search q
Find tokens by symbol, name or contract address across all six chains. Symbol prefix matches rank above name substring matches, and more liquid tokens rank first. If the query is an address FoxD has never indexed, it is resolved on demand from external providers and indexed — `resolved: true` in the response means the token was just learned rather than served from the index. Requires an API key with scope: read.
market.resolve address
Given a bare token address with no chain, return the chain it lives on — the one with the deepest 24h volume if it exists on several. Resolves on demand when the address is not yet indexed. Use this to turn a raw address pasted by a user into the (chain, address) pair every other tool requires. Requires an API key with scope: read.
market.token chain, address
Full detail for one token: price and 5m/1h/6h/24h changes, market cap, FDV, total supply, liquidity, 1h/6h/24h/72h volumes, unique makers, holder count, bonding-curve progress, safety facts (buy/sell tax, honeypot flag, risk level, LP locked) and every native-gas-paired pool. Self-heals — an unindexed or stub token is resolved on demand, so a first call on an unknown address still returns real data. Requires an API key with scope: read.
market.swaps chain, address, limit
The most recent trades against a token, newest first: maker wallet, side, token amount, USD size, execution price, block and timestamp. Use it to read live tape, spot whale entries, or seed a chart. For a continuous stream rather than a snapshot, subscribe to the token channel over the WebSocket feed instead. Requires an API key with scope: read.
market.candles chain, address
Five-minute OHLCV candles for the last 72 hours, oldest first, with USD volume per bucket. Backed by a TimescaleDB continuous aggregate so it is cheap and consistent. Feed this to technical analysis, or call analytics.technicals to have the indicators computed for you. Requires an API key with scope: read.
market.holders chain, address
Top holders for a token with each wallet's share, plus the total holder count and a top-10 concentration figure measured against FREE FLOAT (supply outside the AMM pool), not total supply — measuring against total supply badly understates rug exposure on LP-heavy tokens. Solana reads come from indexed token balances; EVM from a Transfer-log scan. Cached for ten minutes. Requires an API key with scope: read.
market.sparklines tokens
Fetch 72h price series for many tokens at once, keyed by chain:address — up to 800 per call, mixed chains allowed. Use this to hydrate a watchlist or portfolio view without pulling full token summaries for every row. Bound to POST because the key list routinely exceeds a sane URL length. Requires an API key with scope: read.
market.live chain, address
The exact numbers the FoxD trade page shows: price, market cap, liquidity, 1h/6h/24h volume, 1h/24h price change, 24h buy and sell counts, and published socials. Our own indexed data answers first; for tokens we do not cover deeply it falls back to external providers server-side. The `source` field tells you which answered, so two disagreeing figures are explainable rather than mysterious. Prefer this over market.token when you want what a human sees on the terminal right now. Requires an API key with scope: read.
market.tokenMeta tokens
Symbol, name, decimals and logo for many tokens at once, keyed by chain:address — up to 200 per call, mixed chains allowed. Served entirely from our own index with no third party involved, and logos come back as same-origin URLs. Use it to label a list of addresses (a trade history, a portfolio) without making one lookup per row. Requires an API key with scope: read.
analytics.security chain, address
Third-party safety verdict for a token: honeypot flag, buy/sell tax, open-source status, LP locked percentage, normalised risk level (0 safe, higher riskier) and named risk flags. Sourced from honeypot.is on Ethereum/BSC/Base and RugCheck on Solana, refreshed on demand when stale. Providers do not cover Arbitrum or HyperEVM — use analytics.simulateHoneypot there, and to double-check anything before a large buy. Requires an API key with scope: read.
analytics.simulateHoneypot chain, address, platform
Prove a token is actually sellable by simulating a real buy and then a real sell against current chain state across every live venue on the chain. This is ground truth rather than a provider opinion, and it is the ONLY safety source that covers HyperEVM and Arbitrum. Note that `simulated: false` is not a pass — it means no conclusion could be reached (no route, unsupported chain, or a trusted launchpad short-circuit), so check the field before trusting `honeypot: false`. Requires an API key with scope: read.
analytics.token chain, address, sections
Financial analysis computed from FoxD's own swaps and 5m candles — never invented, and null wherever there is not enough history. Four sections: `basics` (24h high/low/VWAP, volume, trade counts, unique traders, average trade size), `technicals` (RSI-14, MACD histogram, Bollinger %B and bandwidth, ATR%, stochastic K, CCI, daily volatility, max drawdown), `quant` (Sharpe, Sortino, Hurst exponent, money-flow index, OBV trend, z-score, volume surge) and `pulse` (a 0-100 heat score with its component breakdown and a multi-timeframe momentum reading). Request only the sections you need. Requires an API key with scope: read.
analytics.curvePrice chain, address
Price for a bonding-curve token that has never traded, derived from the curve's own reserves rather than from swaps. Without this a freshly minted pump.fun or Raydium LaunchLab token has no price at all until its first fill. Covers pump.fun and LaunchLab; Meteora DBC is excluded because it quotes a sqrt-price that is not directly comparable. Returns null for tokens that already trade — use market.token for those. Requires an API key with scope: read.
analytics.score chain, address
The composite FoxScore (1-99) for a token, broken down into the exact weighted terms that produced it: liquidity depth, real participation (unique 24h makers, which is what makes wash-traded tokens score badly), 1h and 24h momentum, buy/sell demand, turnover, holder base, survival age, open-source status, LP lock, and penalties for risk flags and taxes. A confirmed honeypot hard-caps the score at 3. Token summaries carry the headline number; this explains it, so an agent can say WHY something scored badly. Requires an API key with scope: read.
sniper.feed chain, platform, stages, limit, sparklines
New token launches grouped by bonding-curve stage across all six chains and every indexed launchpad (pump.fun, LaunchLab/bonk.fun, Meteora DBC, four.meme, Clanker, Zora, Flaunch, Bankr, alt.fun, Hypurr). Returns four buckets — fresh, bonding, migrated, tracked — each a list of full token summaries carrying bonding progress, liquidity, volume and safety facts. This is the primary sniping surface; use market.newTokens instead when you want raw recency without stage bucketing. Requires an API key with scope: read.
sniper.graduating chain, platform, minProgress, limit
Bonding-curve tokens closest to graduating to an AMM, ranked by how full their curve is. Graduation is the moment liquidity migrates and the price mechanism changes, so it is the single highest-signal window on a launchpad token. Filter by minimum progress to watch only the ones about to tip. Requires an API key with scope: read.
sniper.bondingProgress chain, address
Curve state for one launchpad token: lifecycle stage, curve fill fraction, the deepest pool backing it with its raw reserves, current USD price and liquidity, and the migration timestamp once it graduates. Reserves are returned as raw strings alongside `token0`/`token1` so you can tell which side is which — the ordering follows factory ordering on EVM and base/quote on Solana, and is not normalised. Use this to time an entry precisely rather than polling the whole sniper feed. Requires an API key with scope: read.
sniper.platforms chain
Per-launchpad launch counts and activity over the last 24 hours: how many tokens each pad minted, how many are still bonding, and how many graduated. Use it to see where launch activity has moved before pointing a sniper at a specific pad. Requires an API key with scope: read.
copyIntel.leaderboard chain, sort, category, limit
Ranked wallets from FoxD's own indexed swaps — no third-party data. Each row carries a 0-100 score, 7d realised PnL, win rate, average return, trade and token counts, a behavioural category and a trading style. A quality floor (3-5000 trades and at least $500 volume over 7 days) keeps out both dead wallets and router/bot addresses. Sort by score for all-round quality, by pnl for raw profit, or by winRate for consistency. Requires an API key with scope: read.
copyIntel.categories chain
How many qualifying wallets fall into each behavioural category (smart_money, whale, sniper, defi, mev), optionally for one chain. Use it to size the pool before filtering the leaderboard — if a chain has four smart-money wallets, ranking them is not worth the call. Requires an API key with scope: read.
copyIntel.trader address
Deep analysis of one wallet: lifetime and 24h/7d/30d windows (trades, volume, tokens, realised PnL, win rate), a per-chain breakdown, per-token PnL for its top 100 positions including bags still held, a daily PnL series with a running cumulative curve, and the first page of its trade tape. This is what to read before deciding whether to copy someone. Page further back through the tape with copyIntel.traderTrades. Requires an API key with scope: read.
copyIntel.traderTrades address, before
Paginated trade history for one wallet, newest first, keyset-paginated by timestamp. Pass the oldest timestamp you have seen as `before` to walk backwards. Use this when copyIntel.trader's first page is not enough — for example to reconstruct how a wallet actually entered and exited a specific token. Requires an API key with scope: read.
copyIntel.backtest address, chain, windowDays, ticketUsd, slippageBps
What following this wallet WOULD have returned, replaying its actual past trades at a fixed ticket size with a slippage penalty for arriving late. This answers a different question from the leaderboard: a wallet can be highly profitable and still be a poor thing to copy, because a mirror fills seconds behind and sizes differently. Positions still open at the end of the window are excluded from realised PnL. It is a simulation over past data and says nothing about future performance — present it that way. Requires an API key with scope: read.
trade.chains
Which chains can actually execute a trade right now — the router contracts are deployed and configured on these. A chain FoxD indexes is not necessarily a chain it can trade; check here before quoting, or you will get a chain_not_live error. Also reports whether each chain supports copy trading. Requires an API key with scope: read.
trade.mevProtection
Per-chain MEV protection status: whether a private relay is configured and whether direct public-mempool submission is permitted. Ethereum and BNB Chain submit as private builder bundles so trades never sit exposed where sandwich bots hunt; Base, Arbitrum and HyperEVM use atomic EIP-7702 bundles; Solana broadcasts across multiple RPC paths. Worth checking before sizing up on a volatile pair. Requires an API key with scope: read.
trade.quote chain, tokenIn, tokenOut, amountIn, side, slippageBps, walletId
Price and route a swap WITHOUT executing it: returns the venue, expected output, the slippage-protected minimum that will be enforced on-chain, estimated price impact, and the USD notional the spend cap is measured against. Always quote before executing — it is the only way to see price impact, and a quote costs nothing. A quote is a preview, not a reservation: the pool can move before you execute. Requires an API key with scope: trade:read.
trade.execute chain, tokenIn, tokenOut, amountIn, side, slippageBps, priorityTier, maxPriceImpactBps, walletId
Sign and broadcast a swap from the wallet this API key is bound to. THIS SPENDS REAL FUNDS AND CANNOT BE UNDONE. The trade is quoted first, valued in USD, and refused if it would breach the key's rolling 24-hour spend cap. Slippage is enforced on-chain — if the pool moves beyond tolerance the transaction reverts rather than filling badly. Set maxPriceImpactBps to refuse thin-liquidity fills outright. Check safety with analytics.security AND analytics.simulateHoneypot before buying anything you have not already vetted. Requires an API key with scope: trade:execute.
trade.venues chain
Every DEX and launchpad curve the execution engine can route through on a chain, and whether each is verified live. FoxD verifies a venue with real funds before its buy button goes live, so this is the authoritative list of what can actually be filled against — broader than the set of venues merely indexed for discovery. Requires an API key with scope: read.
trade.feePreview chain, amountIn, priorityTier
What a trade of a given size will cost before you commit. The tip is a FLAT per-chain amount, not a percentage, split into a network incentive (a builder bribe on Ethereum and BNB Chain, a priority-fee budget on the L2s) and FoxD's revenue keep — and the total is capped at 10% of trade value, so a tiny position is never charged more tip than it is worth. `totalTip` is exactly what gets charged; charged equals shown. Also reports the gas reserve held back so a trade can never leave a wallet unable to pay for its next transaction. Requires an API key with scope: trade:read.
wallet.balances walletId
Native coin balances for a wallet on all five EVM chains and Solana — ETH, BNB, HYPE and SOL. Each entry carries both a human amount and the EXACT base units (wei/lamports), because computing a maximum-size trade from the rounded figure will overdraw and fail. A chain whose RPC is unreachable reports zero rather than failing the whole call, so check for suspicious zeroes before concluding a wallet is empty. Requires an API key with scope: wallet:read.
wallet.portfolio walletId
Every holding across all six chains — native coins and tokens, including Solana Token-2022 — priced in USD and sorted by value, plus the total. Also returns `reclaimable`: Solana token accounts that are empty or hold known-worthless dust and whose rent (~0.002 SOL each) can be swept back with wallet.reclaimRent. A token with no known price is NEVER classed as dust, since unpriced does not mean worthless. `degraded: true` means a source failed and the list is incomplete — do not treat it as authoritative. Requires an API key with scope: wallet:read.
wallet.list
Every custodial wallet on the account, one entry each, carrying both its EVM and its Solana address. An account gets one wallet at signup covering both families and may hold up to five. An imported wallet has only the family whose key was imported, so the other address is null. Externally connected sign-in wallets are not included — those are identities, not wallets FoxD can trade from. Requires an API key with scope: wallet:read.
wallet.activity walletId, limit
Unified history for a wallet: trades, transfers, launches, fee claims and rent reclaims, newest first. Built from FoxD's own journal merged with chain-observed transactions, so it includes activity from outside FoxD — a custodial key is exportable, and anything done with it elsewhere still shows up. `gaps` lists chains whose on-chain history could not be read and why, so an empty stretch is never mistaken for genuine inactivity. Requires an API key with scope: wallet:read.
wallet.trades limit
The account's trade history with average-cost realised PnL, win rate, total volume, fees saved and the most-traded tokens. This is FoxD's own journal — every order type lands here (market, limit, DCA, copy and launch dev-buys), so it is the authoritative record of what the account actually did, unlike a wallet's raw chain history. Requires an API key with scope: wallet:read.
wallet.gasEstimate chain, units
How much native coin to hold back on a chain so a transaction can pay for itself. Subtract this before sending a wallet's entire balance — sending the full amount leaves nothing for gas and the transaction simply fails. Reflects live network conditions rather than a fixed guess. Requires an API key with scope: wallet:read.
wallet.send chain, to, amount, max, walletId
Transfer native coin from a custodial wallet to any address. THIS MOVES REAL FUNDS AND CANNOT BE UNDONE — there is no recall, and a wrong address is a permanent loss. Counts against the key's rolling 24h spend cap. Call wallet.gasEstimate first and hold that much back, or a full-balance send fails for gas; or set `max` and the wallet is swept to near-zero with the exact gas withheld for you. FoxD charges no withdrawal fee — you pay only network cost. Requires an API key with scope: wallet:write.
wallet.reclaimRent preview, account, walletId
Every SPL token fully sold leaves an empty account still holding its ~0.002 SOL rent deposit. This closes them and returns the SOL. Call with `preview: true` first to see the count and amount without sending anything. Only truly empty or known-worthless-dust accounts are ever closed — a token with no known price is never burned, because unpriced does not mean worthless. Accounts frozen by their token cannot be reclaimed. EVM has no equivalent: ERC-20 balances live in the contract's mapping, so holding a token costs nothing there. Requires an API key with scope: wallet:write.
wallet.create label
Provision an additional custodial wallet. Each carries both an EVM and a Solana address, and an account may hold up to five. Useful for separating strategies — a wallet per agent means one misbehaving bot cannot reach the others, since a key binds to exactly one wallet. Requires an API key with scope: wallet:write.
wallet.update walletId, label, isDefault
Change a wallet's label, or make it the account default. The default is what tools use when no walletId is given. An externally connected sign-in wallet can never be made default — FoxD holds no key for it and cannot trade from it. Requires an API key with scope: wallet:write.
wallet.delete walletId
Remove a wallet from the account. EXPORT THE PRIVATE KEY FIRST IF IT HOLDS ANYTHING — deletion removes FoxD's copy of the key, and any funds left in it become unreachable. There is no recovery. Requires an API key with scope: wallet:write.
wallet.import privateKey, label
Bring a wallet you already control into FoxD so it can be traded from. Accepts a 0x-prefixed EVM key or a base58 Solana key; the key is sealed with AES-256-GCM at rest. Importing hands FoxD the ability to sign for that wallet — do it only with a key whose funds you are willing to trade through this platform, never with a cold-storage or multi-purpose key. Requires an API key with scope: wallet:write.
wallet.export walletId, chainFamily
Return the private key of a FoxD-provisioned wallet, so you can move it elsewhere — your funds are yours and are never locked in. UNDERSTAND WHAT THIS MEANS FOR AN API KEY: a private key is not bounded by scopes, spend caps or revocation. Anyone holding it controls that wallet permanently, and revoking the API key afterwards changes nothing. Only call this to hand the key to its human owner, never to store it, log it, or pass it to another service. Imported wallets cannot be exported this way. Requires an API key with scope: wallet:write.
automation.orders status, limit
Automated orders on the account — limit, DCA, take-profit, stop-loss and trailing stops — with their triggers, progress and state. Defaults to what is currently armed; pass `status` to see what filled, what you cancelled, or what was parked after repeatedly failing to execute. This is how you find out an order fired without polling a balance. For a trailing order, `peakPriceUsd` is the best price seen since it armed, which is what the stop is measured down from. Requires an API key with scope: automation:read.
automation.placeLimit chain, tokenAddress, amount, slippageBps, priorityTier, walletId, side, targetPriceUsd
Buy when the price falls to a target, or sell when it rises to one. The order stays armed until it fires or you cancel it. Triggers evaluate against FoxD's live market price, which is fresh but not instantaneous — in a fast market the fill can differ from the trigger price, bounded by this order's slippage tolerance. Requires an API key with scope: automation:write.
automation.placeTakeProfit chain, tokenAddress, amount, slippageBps, priorityTier, walletId, targetPriceUsd
Sell a held position once the price RISES to a target, locking in a gain. Fires once, then completes. Triggers evaluate against FoxD's live market price, which is fresh but not instantaneous — in a fast market the fill can differ from the trigger price, bounded by this order's slippage tolerance. Requires an API key with scope: automation:write.
automation.placeStopLoss chain, tokenAddress, amount, slippageBps, priorityTier, walletId, targetPriceUsd
Sell a held position once the price FALLS to a target, capping a loss. Fires once, then completes. On an illiquid token a stop can fill well below the trigger — size the slippage tolerance accordingly. Triggers evaluate against FoxD's live market price, which is fresh but not instantaneous — in a fast market the fill can differ from the trigger price, bounded by this order's slippage tolerance. Requires an API key with scope: automation:write.
automation.placeTrailingStop chain, tokenAddress, amount, slippageBps, priorityTier, walletId, trailPercent, activateAboveUsd
Follow the price up and sell after it retraces by a set percentage from its best. Leaving `activateAboveUsd` unset trails from the moment it is placed — a trailing STOP-LOSS. Setting it makes the order wait until that price is reached before it starts trailing, which is a trailing TAKE-PROFIT: it protects a gain rather than capping a loss. A wider trail survives normal volatility but gives back more of the move. Triggers evaluate against FoxD's live market price, which is fresh but not instantaneous — in a fast market the fill can differ from the trigger price, bounded by this order's slippage tolerance. Requires an API key with scope: automation:write.
automation.placeDca chain, tokenAddress, amount, slippageBps, priorityTier, walletId, side, intervalSec, runsTotal
Buy or sell a fixed amount on a repeating interval for a set number of runs, averaging the entry rather than timing it. Each run executes independently, so a single failure does not cancel the schedule. Requires an API key with scope: automation:write.
automation.cancelOrder orderId
Disarm an active order so it never fires. An order already mid-execution cannot be pulled back — cancelling is not a way to undo a fill that has started. Requires an API key with scope: automation:write.
automation.strategies
Saved trading strategies on the account with their rules and status (draft, active, paused or stopped), plus how many times each has triggered. A strategy is a named set of conditions; only an active one runs. Requires an API key with scope: automation:read.
automation.createStrategy name, description, rules
Save a strategy. Supply `rules` directly, or give only a plain-language `description` and it will be parsed into rules for you. Strategies are created as DRAFT and do nothing until you activate them with automation.updateStrategy — so a misparsed description cannot trade on its own. Requires an API key with scope: automation:write.
automation.updateStrategy strategyId, name, description, rules, status
Change a strategy's name, description or rules, or move it between draft, active, paused and stopped. Setting status to active is what actually lets it trade — review the parsed rules before you do. Requires an API key with scope: automation:write.
automation.deleteStrategy strategyId
Permanently remove a strategy. It stops immediately and cannot be restored. Requires an API key with scope: automation:write.
copyExec.subscriptions
Every wallet this account is currently mirroring, with the sizing rule, caps, filters and whether each is enabled. A disabled subscription is retained but does not trade. `follower` is the wallet each one spends from — an unfunded follower means nothing will mirror, however correct the configuration looks. Requires an API key with scope: copy:read.
copyExec.subscribe leader, chain, sizing, walletId, label, maxPerTradeIn, minLeaderAmountIn, copySides, allowTokens, denyTokens, slippageBps, tier
Mirror a leader wallet's trades from your own funds. EVERY MIRRORED TRADE SPENDS REAL MONEY, automatically, for as long as this stays enabled — this is the single most consequential thing an agent can set up here, because it keeps trading without further approval. Vet the leader with copyIntel.trader and copyIntel.backtest first, and set maxPerTradeIn. Mirroring executes custodially from the returned `follower` wallet the moment the leader trades, so fund that address; mirrors land a few seconds behind, and a fast-moving entry can fill at a different price than they got. One subscription per (leader, chain): subscribing twice returns 409. Requires an API key with scope: copy:write.
copyExec.updateSubscription subscriptionId, enabled, sizing, label, maxPerTradeIn, minLeaderAmountIn, copySides, allowTokens, denyTokens, slippageBps, tier
Change sizing, caps, filters or slippage — or set `enabled: false` to stop mirroring immediately while keeping the configuration. Disabling is the fast way to halt a leader who has started behaving badly. Only the fields you pass are changed. Requires an API key with scope: copy:write.
copyExec.unsubscribe subscriptionId
Remove a copy subscription entirely. Mirroring stops immediately; positions already opened by it are untouched and remain yours to manage. Requires an API key with scope: copy:write.
launch.pads
Every launchpad FoxD can deploy to, which chain each runs on, its base cost, and whether it is currently available. Availability is not static — a pad whose API or config is unhealthy is reported unavailable rather than failing mid-deploy, so check here before quoting. Requires an API key with scope: read.
launch.quote pads, devBuys
Itemised cost of deploying across a chosen set of launchpads, including optional per-pad dev buys, plus whether the bound wallet can actually afford it. Quote before launching — a deploy that runs out of funds partway leaves some variants live and others not. Requires an API key with scope: launch:read.
launch.create name, symbol, description, imageBase64, imageMimeType, pads, devBuys, options, twitter, telegram, discord, website
Deploy ONE token — same name, symbol, image and socials — across every launchpad you select, and get back a single id that unifies the variants. This SPENDS REAL FUNDS from the bound wallet on every pad chosen, and a deployed token cannot be undone, renamed or un-taxed, so call launch.quote first and check the caller actually intends it. Deployment is asynchronous: this returns as soon as the launch is recorded and the metadata is pinned, with the variants queued; poll launch.variants to watch them go live. The image is required and passed as base64 — there is no URL fetch. Every chain the launch touches is checked for funds in its own currency before anything is written, so a Base pad selected with an empty ETH balance is refused up front rather than failing after the metadata is permanently pinned. Requires an API key with scope: launch:write.
launch.variants multiverseId
Every deployed variant of a multiverse launch with its mint, pool, per-pad price, liquidity, volume and momentum — the 'which one is winning' view. A variant still deploying shows as pending; one that failed shows why, so a partial launch is visible rather than silent. Requires an API key with scope: read.
launch.studio
Every launch this account created, with combined metrics and the creator fees currently claimable — split by whether they sit in a per-creator vault or a per-pool account, because those are claimed differently. Also reports lifetime claims per chain. This is the creator dashboard. Requires an API key with scope: launch:read.
launch.claimFees platform, variantId
Sweep creator trading fees from your launches into your wallet. Claim one launchpad or variant, or everything by omitting both. THIS SENDS TRANSACTIONS. Each pad is attempted independently and reported separately, so one failing pad never blocks the rest — read the per-attempt outcomes rather than assuming overall success. Clanker settles in WETH on Base from your EVM wallet; the Solana pads settle in SOL. Requires an API key with scope: launch:write.
launch.explore limit, status, symbol
Every token launched through the FoxD Meme Multiverse, newest first — not just your own. Each entry carries the launch identity, its creator wallet, status and how many variants went live. This is the discovery surface: use it to see what is being launched right now, then launch.variants to compare a specific launch's pads against each other. Requires an API key with scope: read.
launch.migrate multiverseId, fromMint, toMint, amountIn, slippageBps
Sell one variant of a multiverse launch and buy another — the 'ride the winner' move, executed as two legs through the normal trade path. Both mints must be LIVE variants of the SAME launch, so this can never become a general-purpose swap. IT IS NOT ATOMIC: the sell can succeed and the buy fail, leaving you in native coin rather than either token. The response reports which legs completed so a partial migration can be retried from where it stopped rather than repeated from the start. Requires an API key with scope: trade:execute.
alerts.list
Every webhook on this account with its event, filters, delivery health and last error. A subscription that fails ten times consecutively is paused rather than retried forever — check `status` before assuming events are still flowing. Requires an API key with scope: read:account.
alerts.subscribe url, event, filters
Be POSTed to when an event happens instead of polling for it. The URL must be https and publicly routable — private, loopback and link-local addresses are refused, because a webhook makes OUR servers call an address YOU choose. Returns a signing secret shown ONCE: store it, and verify every delivery's signature before trusting the payload. Filters narrow the firehose; without them, new_token means every launch on every chain. Each delivery is a POST of {eventId, event, timestamp, data} with an x-foxd-signature header — HMAC-SHA256 of "<timestamp>.<body>" under your secret. Delivery is at-least-once, so deduplicate on eventId. Ten consecutive failures pause the subscription.
Requires an API key with scope: read:account.alerts.unsubscribe webhookId
Remove a webhook. Deliveries stop immediately and the signing secret is destroyed. Requires an API key with scope: read:account.
account.whoami
Who this API key acts for: the user id, display name, which sign-in provider created the account, the key's granted scopes and tier, the wallet it is bound to, and its remaining spend allowance. Call this first from any agent to confirm what it is actually permitted to do before attempting something it will be refused. Requires an API key with scope: read:account.
account.wallets
The custodial wallets provisioned for this account — one entry per wallet, each carrying its EVM address and its Solana address. Signup provisions one wallet covering both; an account may hold up to five. These are the addresses to fund: an agent that just signed up should read this, fund the returned addresses, then trade. Use wallet.balances for what is actually in them. Requires an API key with scope: wallet:read.
account.usage
Month-to-date calls and weighted cost units for this key against its tier limits, plus the per-minute rate ceiling. Everything is free during early access, but quotas are real — read this to pace a long-running agent rather than discovering the limit by being refused mid-task. Requires an API key with scope: read:account.
account.preferences
Account-level settings shared with the web terminal: display name, avatar, theme, language, default slippage and whether small balances are hidden. An agent reading defaultSlippage here will trade the way the user configured the UI to trade, instead of inventing its own tolerance. Requires an API key with scope: read:account.
account.updatePreferences displayName, theme, language, defaultSlippage, hideSmallBalances
Change account settings. Only the supplied fields are modified, and changes appear immediately in the web terminal since both read the same row. Restricted to a fixed allowlist of harmless display and trading-default fields — it can never touch scopes, wallets or keys. Requires an API key with scope: read:account.
keys.list
Every active and paused key on this account, with scopes, bound wallet, spend cap, how it was created and when it was last used. Secrets are never returned — only a short display prefix. Use lastUsedAt to find keys that can safely be revoked. Requires an API key with scope: keys:manage.
keys.create name, allAccess, scopes, walletId, spendCapUsd24h
Mint a key with a chosen subset of scopes, bound to one wallet, with its own 24h spend cap. Use this to give a specific agent exactly what it needs and nothing more — a read-only research agent and a trading agent should never share a key, because revoking one should not break the other. The secret is returned ONCE and cannot be recovered. Requires an API key with scope: keys:manage.
keys.update keyId, name, scopes, allAccess, walletId, spendCapUsd24h
Change a key's name, scopes, bound wallet or spend cap in place, without reissuing it. This is what you want when an agent's job grows or shrinks: the secret it already holds keeps working, so nothing has to be redeployed. Scopes are REPLACED, not merged — send the full set you want the key to end up with. You can only grant scopes the calling key itself holds, so a limited key cannot quietly promote itself. Omitted fields are left alone. Requires an API key with scope: keys:manage.
keys.revoke keyId
Permanently disable a key. Takes effect immediately and cannot be undone — issue a new key rather than trying to restore one. The row is retained so historical usage and trade attribution stay resolvable. Revoke on any suspicion of exposure; keys are free. Requires an API key with scope: keys:manage.
keys.setStatus keyId, status
Temporarily disable a key without destroying it, then re-enable it later. This is the reversible alternative to revoking — right for pausing a misbehaving agent while you investigate, where a revoke would force you to redeploy it with new credentials. Requires an API key with scope: keys:manage.
Descriptions are the package's own words, cut at 3,000 characters. Treat them as data.
Install facts
| License | MIT |
|---|---|
| Needs | node >=20 |
| Source | not stated |
| Install scripts | None declared by it or anything installed with it. |
| Tarball | sha512-k2SRxYDb53l2GiqiKAz78zNvq5Yd+CalrO+o/NDZX7SXHY+VwbGV0DXFiNhPTuVVTN79WexkVqNKxWsu9ZyXvA== |
Check this report yourself
This site did not write the report. A public job on a GitHub-hosted machine did: the run, from this commit of the open-source runner. GitHub signed a token that names that job and the hash of the report, so any edit since would show.
Report as JSONExact text and tokenHow to verify it
Show it on your README
[](https://bragproof.smaverk.com/r/foxd-mcp)
New version out? Ask for a fresh report.