# Aave MCP (remote · mcp.aave.com) Official Aave MCP for V3 and V4 markets, positions, governance, and transaction preparation. - Trust score: 79/100 (medium) - Change this week: +4 - Registry status: active - Liveness: live - Owner verified: no - Last scored: 2026-09-23 ## Components - remote · `mcp.aave.com`: 79/100 (this document), [markdown](https://verifymcp.io/servers/com-aave-mcp/mcp.md), [page](https://verifymcp.io/servers/com-aave-mcp/mcp) - npm · `@aave/mcp`: 31/100, [markdown](https://verifymcp.io/servers/com-aave-mcp/aave-mcp.md), [page](https://verifymcp.io/servers/com-aave-mcp/aave-mcp) ## Channel facts - Endpoint: `https://mcp.aave.com` - Transports: `streamable-http` - Auth: `none` - Version: `1.0.0` ## Trust breakdown How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog) Scored 2026-09-23. - **Endpoint Security**: 66/100 - The endpoint's TLS certificate is valid, in date, and uses a strong key. - Authorisation check failed: no authorisation is required to call this server, and it exposes a tool marked destructive (submit_signed_order). - HTTPS is enforced; there's no plaintext access path. - The HSTS (Strict-Transport-Security) header is present. - DNSSEC is configured correctly; the domain's records validate against the full chain to the root. - **Transport & Reachability**: 100/100 - Verified streamable-http transport via a live MCP handshake. - **Schema Quality & AI Usability**: 80/100 - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name). - AI-judged instruction clarity (excellent). - Context-footprint check failed: tool/resource definitions use about 15076 tokens (~212/item across 71 items; 53 tools + 18 resources), over budget; trim descriptions and params. - Usage-examples check failed: none of the tools include examples. - **Stability & Change Management**: 70/100 - Stability observed for 21 of 30 days with no destabilising changes; credit accrues until the full window elapses. - **Tool Coverage**: 100/100 - 100% of tools have a non-trivial description (not blank, and not just the tool's name). - 100% of tool parameters carry a description. - **Tool Safety**: 100/100 - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured. - We read all 53 captured tool definition(s), and no name or description among them implies an irreversible operation. - An AI judge read all 55 captured unit(s) of tool text and found none that tries to manipulate the model reading it. - **Capabilities**: 100/100 - Implements a current MCP spec version (2026-07-28). ## Install ### How do I install the Aave MCP server? Aave MCP is a hosted endpoint at https://mcp.aave.com/, so there is nothing to install locally. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation. ### Claude ```bash claude mcp add --transport http com-aave-mcp 'https://mcp.aave.com/' ``` ### Cursor ```json { "mcpServers": { "com-aave-mcp": { "url": "https://mcp.aave.com/" } } } ``` ### VS Code ```json { "servers": { "com-aave-mcp": { "type": "http", "url": "https://mcp.aave.com/" } } } ``` ### Codex ```toml [mcp_servers.com-aave-mcp] url = "https://mcp.aave.com/" ``` ### opencode ```json { "$schema": "https://opencode.ai/config.json", "mcp": { "com-aave-mcp": { "type": "remote", "url": "https://mcp.aave.com/", "enabled": true } } } ``` ### OpenClaw ```bash openclaw mcp add com-aave-mcp --url 'https://mcp.aave.com/' --transport streamable-http ``` ### Hermes ```yaml mcp_servers: com-aave-mcp: url: "https://mcp.aave.com/" ``` ### Netclaw ```json { "McpServers": { "com-aave-mcp": { "Transport": "http", "Url": "https://mcp.aave.com/" } } } ``` ### Vellum ```bash assistant mcp add com-aave-mcp -t streamable-http -u 'https://mcp.aave.com/' ``` ### Other ```json { "mcpServers": { "com-aave-mcp": { "type": "http", "url": "https://mcp.aave.com/" } } } ``` The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs. ## Changelog Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims. ### 2026-09-23 (score 79, +1) No change was recorded against any check on this day. Stability & Change Management went from 67 to 70. That category is still filling its 30-day observation window: 20 days of observed history at the previous scan, 21 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-09-21 (score 78, +1) No change was recorded against any check on this day. Stability & Change Management went from 60 to 63. That category is still filling its 30-day observation window: 18 days of observed history at the previous scan, 19 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-09-19 (score 77, +1) No change was recorded against any check on this day. Stability & Change Management went from 53 to 57. That category is still filling its 30-day observation window: 16 days of observed history at the previous scan, 17 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-09-18 (score 76, 0) - [security] Tool “get_market_history” rewrote its description, which is the text the model reads - [security] Tool “get_spoke_configs” rewrote its description, which is the text the model reads - [cosmetic] “get_vaults” added an optional parameter “cursor” - [cosmetic] “get_spoke_configs” reworded the description of “spokeId” - [cosmetic] “get_vault_activity” reworded the description of “window” - [cosmetic] “get_spoke_configs” reworded the description of “hubId” ### 2026-09-17 (score 76, +1) No change was recorded against any check on this day. Stability & Change Management went from 47 to 50. That category is still filling its 30-day observation window: 14 days of observed history at the previous scan, 15 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-09-16 (score 75, 0) - [security] The server rewrote its instructions, which are the text every model session reads - [security] Tool “get_aave_guide” rewrote its description, which is the text the model reads - [security] Tool “get_sgho_vault” rewrote its description, which is the text the model reads - [security] Tool “get_user_summary” rewrote its description, which is the text the model reads - [security] Tool “prepare_stkgho_migrate” rewrote its description, which is the text the model reads - [functional regression] Schema quality: 11508 → 14941 - [functional] New prompt “portfolio” - [functional] New resource “aave-guide-safety-module” - [functional] New resource “aave-guide-portfolio” - [functional] New tool “get_asset_history” - [functional] New tool “get_interest_rate_model” - [functional] New tool “get_market_history” - [functional] New tool “get_multichain_asset” - [functional] New tool “get_reserve_holders” - [functional] New tool “get_risk_premium” - [functional] New tool “get_safety_module” - [functional] New tool “get_spoke_configs” - [functional] New tool “get_vault_activity” - [functional] New tool “get_vault_preview” - [functional] New tool “get_vaults” - [functional] New tool “get_wallet_balances” - [functional] New tool “prepare_vault_action” ### 2026-09-15 (score 75, +1) No change was recorded against any check on this day. Stability & Change Management went from 40 to 43. That category is still filling its 30-day observation window: 12 days of observed history at the previous scan, 13 at this one. The score rises as the window fills, whether or not the server changes. ### 2026-09-13 (score 74, +1) No change was recorded against any check on this day. Stability & Change Management went from 33 to 37. That category is still filling its 30-day observation window: 10 days of observed history at the previous scan, 11 at this one. The score rises as the window fills, whether or not the server changes. ## MCP tools (53) ### `get_chains` (~111 tokens) Get supported chains List chains Aave supports, on v3, v4, or both. A chain carrying 'notServed': true is one this API holds no market on: every market, position and rewards read comes back empty for it, and that emptiness says nothing about the chain itself. Its gateway addresses read 0x0 for the same reason, so do not send a transaction to one. Every chain without the flag has markets to read. Input parameters: - `version` (string): Protocol version (default all). ### `get_markets` (~562 tokens) Get markets / reserves List Aave markets/reserves (symbol, APY, caps, liquidity) on v3, v4, or both. Chain coverage: omit chainId and every response states the chains it read as 'chainsCovered', plus 'chainsNotCovered' when it read fewer than all of them. 'chainsNotServed', when present, lists chains this API holds no market on: Aave supports the chain, there is nothing here to read, so an empty result for one of them is not an answer about that chain, and asking again returns the same nothing. v4 always covers every chain it serves; v3 covers every chain when you pass 'symbols' and Ethereum alone when you do not, because an unfiltered 21-chain listing is hundreds of kilobytes. So to compare one asset across chains, pass symbols and no chainId. Every rate is a percent ('3.32' means 3.32%), named with a Pct suffix. Do not rank on APY alone: 'suppliable'/'borrowable' (v4), 'availableLiquidity' and 'totalSuppliedUsd' (v3) say whether a rate can actually be entered, and isFrozen / isPaused / supplyCapReached / borrowCapReached appear only when set, so a reserve with none of them is not flagged. Both versions return one row per reserve under 'reserves', carrying supplyApyPct and borrowApyPct together plus canSupply / canBorrow for the sides that are actually open, so the two versions read the same way. On v4 one asset appears once per spoke, so rows are labelled with 'spoke' and 'hub'; spokes sharing a hub share its rate, so identical APYs across rows are expected rather than duplicates. 'canSwapFrom' (v4) says whether that token can be the sell side of a token swap, which is the per-reserve form of what get_swappable_tokens answers per chain. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `symbols` (array): Optional token symbols to return, e.g. ['USDC','GHO']. Case-insensitive. Strongly preferred when you know the asset: a chain has ~70 reserves and listing them all is a large response. - `user` (string): Optional wallet address (0x, 40 hex). Adds walletBalance, youCanSupply and youCanBorrow to every row, for this wallet, in the same call. Pass it whenever the question is about what someone can actual… - `version` (string): Protocol version (default all). ### `get_emode_categories` (~249 tokens) Get v3 eMode categories v3 only. eMode groups correlated assets (ETH-correlated, stablecoins) so they borrow against each other at a higher LTV, raising borrowing power at the cost of restricting which assets the position may hold. Returns each market's categories with the 'categoryId' that prepare_set_emode takes, its LTV and liquidation thresholds, and which symbols are in it. Narrow with 'symbols' (e.g. ['USDC']) to get just the categories that asset is in - unfiltered this is every category on every chain. Markets with no matching category are left out entirely. v4 has no eMode: it replaces it with risk premium and dynamic config. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `symbols` (array): Optional token symbols to return, e.g. ['USDC','GHO']. Case-insensitive. Strongly preferred when you know the asset: a chain has ~70 reserves and listing them all is a large response. - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `get_user_positions` (~261 tokens) Get user positions A wallet's Aave positions (supplies and borrows) on v3, v4, or both. Omit chainId and both versions cover every chain they serve, stating which under 'chainsCovered'; there is no need to loop over chains yourself. 'chainsNotServed', when present, lists chains this API holds no market on: Aave supports the chain, there is nothing here to read, so an empty result for one of them is not an answer about that chain, and asking again returns the same nothing. v4 returns 'positions', each with its own health factor; v3 returns 'supplies' and 'borrows'. These are Aave-side positions only: a wallet's own token balances are not here, and an empty result means nothing is supplied or borrowed rather than that the wallet is empty. For what it holds, pass 'user' to get_markets, which returns walletBalance per reserve. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Protocol version (default all). ### `get_position_items` (~138 tokens) Get individual position items (v4) v4 only. The individual supplies or borrows inside one v4 spoke: principal, accrued interest and what is withdrawable right now, per item, which the per-spoke aggregate from get_user_positions cannot show. Get 'spokeId' from get_user_positions. Input parameters: - `side` (string): Which side to list (default supply). - `spokeId` (string, required): spokeId from get_user_positions (not the spoke address). - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_user_summary` (~370 tokens) Get user summary / health factor A wallet's aggregate position and health factor. Omit chainId and both versions cover every chain they serve, stating which under 'chainsCovered'; there is no need to loop over chains yourself. 'chainsNotServed', when present, lists chains this API holds no market on: Aave supports the chain, there is nothing here to read, so an empty result for one of them is not an answer about that chain, and asking again returns the same nothing. Ask for one of those chains on its own and the reply is the scope alone, with no totals: a zero would read as a real balance. v4 returns a cross-chain summary (lowest health factor, totals, net APY). v3 has no cross-market aggregate, so it returns one entry per market the wallet actually holds a position in, each with its own health factor; 'marketsWithPosition': 0 means the wallet holds no v3 position anywhere, not that the read was narrow or that the wallet is empty: token balances are not here. For what it holds, pass 'user' to get_markets, which returns walletBalance per reserve. This is the lending position and only that. A wallet's Aave holdings also include its Safety Module stake (get_safety_module) and Savings GHO (get_sgho_vault), neither of which is a reserve and neither of which appears in any figure here, so a portfolio answer built from this alone understates what the wallet holds. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Protocol version (default all). ### `get_transaction_processed` (~168 tokens) Has Aave seen my transaction yet? v4 only. After sending a transaction, ask whether the Aave API has processed it. Use this instead of sleeping or polling an RPC: the API lags the chain, and prepare_action will not issue a dependent follow-up (the supply after an approval, the withdraw after a repay) until it has caught up. Pass the txHash you sent plus the 'operations' array that came back on that transaction. Poll until processed is true, then continue. Input parameters: - `operations` (array, required): The 'operations' from the transaction the server built, e.g. ['SPOKE_SUPPLY']. - `txHash` (string, required): Transaction hash you broadcast (0x...). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_reserve_details` (~403 tokens) Get reserve details Deep detail for a single reserve (rates, caps, risk parameters, utilisation, the interest-rate curve, token addresses) on v3 or v4. Every rate and percentage is a percent ('3.32' means 3.32%), named with a Pct suffix. On v4, utilisation and the rate curve are hub-wide, shared by every spoke drawing on the asset. Two v4 fields whose meaning the numbers do not give away: 'collateralFactorPct' is the liquidation boundary itself, because v4 has no separate liquidation threshold - do not look for one - so a position drawn to its full borrow limit sits at health factor 1.0; and 'liquidityFeePct' is the protocol's share of borrower interest, already netted out of supplyApyPct, so treating it as a user cost double-counts it (supplyApy = borrowApy x utilisation x (1 - liquidityFee), if you want to check). This is also where the asset's price lives: 'priceUsd', with 'priceSource' naming who priced it, and no listing carries either - get_markets quotes no price at all, so what an asset is worth on Aave is answered here. Use get_markets first to get the reserveId (v4) or market+token (v3). Input parameters: - `chainId` (number): v3 only: chain id (positive integer). - `market` (string): v3 only: market pool address (from get_markets). - `reserveId` (string): v4 only: the reserveId from get_markets, copied verbatim. - `token` (string): v3 only: underlying token address. - `version` (string): Optional: inferred from the reserve selector ('reserveId' is v4, 'market'+'token'+'chainId' is v3). Send it to be explicit, or if you somehow set both. ### `get_apy_history` (~309 tokens) Get APY history Historical supply or borrow APY for a reserve over time, on v3 or v4. Returns a time-series of {date, apyPct}, where apyPct is a percent ('3.32' means 3.32%). Each point is an average over its sampling interval, and the interval widens with the window (hourly for 'day', coarser above that), so a series is not directly comparable to the instantaneous rate from get_reserve_details. Use get_markets first for the reserveId (v4) or market+token (v3). When comparing v3 against v4 for the same asset (a migration decision), read the history on both sides: a spot-rate gap can be one side's momentary spike. Input parameters: - `chainId` (number): v3 only: chain id (positive integer). - `market` (string): v3 only: market pool address (from get_markets). - `reserveId` (string): v4 only: the reserveId from get_markets, copied verbatim. - `side` (string): Which rate (default supply). - `token` (string): v3 only: underlying token address. - `version` (string): Optional: inferred from the reserve selector ('reserveId' is v4, 'market'+'token'+'chainId' is v3). Send it to be explicit, or if you somehow set both. - `window` (string): Time window (default week). ### `get_user_activity` (~265 tokens) Get user activity / transaction history A wallet's Aave transaction history (supply, borrow, repay, withdraw, collateral changes) on v3, v4, or both. Most recent first. v4 covers every chain; v3 reads one market on one chain at a time, because upstream paginates per market, so it states its scope as 'chainsCovered' and 'market' and you pass chainId to read another. Use get_user_positions, not this, to find out where a wallet holds anything. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `cursor` (string): Pagination cursor: copy 'pageInfo.next' from a previous call verbatim. It is opaque and cannot be edited, shortened or rebuilt - a value that was changed comes back as a base64 parse failure. Require… - `market` (string): v3 only, optional: market pool address (defaults to the chain's main market). - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Protocol version (default all). ### `get_protocol_history` (~141 tokens) Protocol totals over time (v4) v4 only. Aave v4 protocol-wide history: market size, total borrows and available liquidity per sample over a window. Omit chainId to cover every supported v4 chain. Use this for 'is the protocol growing', not for one wallet (get_user_summary_history) or one reserve (get_apy_history). Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `version` (string): Optional, and only 'v4': this tool exists on v4 only. - `window` (string): Time window (default week). ### `get_user_summary_history` (~156 tokens) A wallet's position over time (v4) v4 only. How one wallet's net worth, supplied, debt and health factor moved over a window on Aave v4. Use it to answer 'how has my position changed' or to see how close to liquidation a wallet has been. Omit chainId to cover every supported v4 chain. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. - `window` (string): Time window (default week). ### `get_hubs` (~127 tokens) Get v4 hubs v4 only. Aave v4 hubs: a hub holds the liquidity and global accounting that user-facing spokes draw from. Returns each hub's hubId, address, chain and totals. Pass a hubId to get_hub_assets for the assets it carries. Omit chainId to cover every supported v4 chain. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_hub_assets` (~117 tokens) Get the assets a v4 hub carries v4 only. The assets held by one Aave v4 hub, with hub-wide supplied and borrowed totals, rates and utilisation. Get hubId from get_hubs (or the hubAddress on get_reserve_details). This is hub-level accounting; for what a user can actually supply or borrow, use get_markets, which is per spoke. Input parameters: - `hubId` (string, required): hubId from get_hubs. - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_market_history` (~289 tokens) Market totals over time (v4) v4 only. Deposits, borrows, available liquidity and utilisation over a window, for one hub or one spoke. Pick the layer deliberately: a hub is the liquidity itself and answers 'is this market growing', a spoke is one venue onto that liquidity and answers 'is this venue being used'. Ids come from get_reserve_details, which returns both 'hubId' and 'spokeId' for a reserve, or from get_hubs ('hubId') and get_user_positions ('spokeId' for a spoke the wallet already holds). The two are not interchangeable and the wrong one is rejected. To judge a venue before entering it, take the ids from get_reserve_details rather than from a position that does not exist yet. For the protocol as a whole rather than one market use get_protocol_history, for one reserve's rate use get_apy_history, and for one asset across every reserve holding it use get_asset_history. Input parameters: - `id` (string, required): The hubId (scope 'hub') or spokeId (scope 'spoke'). Not an address. - `scope` (string, required): Which layer 'id' names: a hub's liquidity, or one spoke onto it. - `version` (string): Optional, and only 'v4': this tool exists on v4 only. - `window` (string): How far back to go (default week). ### `get_asset_history` (~280 tokens) One asset's price, supply or borrows over time (v4) v4 only. One asset across every v4 reserve that holds it: 'price' is its oracle price per sample, 'supply' and 'borrow' are the amounts plus the APY spread across those reserves (averageApyPct with the highest and lowest at that moment). Not the same question as get_apy_history, which is one reserve's own rate - an asset sits in several reserves at once, so a wide gap between highestApyPct and lowestApyPct is the point, not noise. The asset is named by its token address and chain, and get_markets does NOT return a token address, so do not look for one there: get_multichain_asset takes a symbol and returns the token address per chain in one call, and get_hub_assets returns it per hub asset. Input parameters: - `chainId` (number, required): Chain the token is on, e.g. 1 for Ethereum. Required: the same symbol exists on several. - `metric` (string, required): Which series: oracle price, supplied amount, or borrowed amount. - `token` (string, required): Token address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. - `window` (string): How far back to go (default week). ### `get_reserve_holders` (~232 tokens) Largest suppliers or borrowers of a reserve (v4) v4 only. The biggest wallets supplying or borrowing one v4 reserve, largest first, for questions about concentration: whether one address is most of a reserve's liquidity, or most of its debt. The token is named once for the whole list rather than repeated per row. 'reserveId' comes from a get_markets row or a get_position_items item. An empty list means this reserve has no holders indexed on that side, which is common for the borrow side of an asset few people borrow, and is not an error. Input parameters: - `cursor` (string): 'nextCursor' from a previous call, for the next page. Omit the argument entirely for the first page. - `limit` (number): How many wallets to return. Only 10 or 50; default 10. - `reserveId` (string, required): reserveId from get_markets or get_position_items. - `side` (string): Rank the wallets supplying it, or the ones borrowing it (default supply). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_interest_rate_model` (~170 tokens) The interest rate curve for an asset (v4) v4 only. The rate curve for one hub asset: borrow and supply APY at each level of utilisation, as points along the curve the protocol actually uses. This explains a rate rather than reporting it - where the kink sits, how steeply borrowing gets more expensive past it, and what the rate would become if utilisation moved. Use it before a large supply or borrow that would shift utilisation, because the current rate from get_reserve_details is one point on this curve and not what a large position would pay. 'hubAssetId' comes from get_hub_assets. Input parameters: - `hubAssetId` (string, required): hubAssetId from get_hub_assets (not a token address). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_wallet_balances` (~226 tokens) What a wallet holds that Aave accepts (v4) v4 only. The tokens a wallet actually holds that v4 can take, per token and summed across chains, with 'perChain' saying where each sits, plus the best supply APY and lowest borrow APY the protocol currently offers on that token. This is the wallet, not a position: nothing here is supplied or borrowed yet, and the rates alongside a balance are what Aave would pay or charge, not what the wallet is earning. Start a 'what should I do with what I have' question here and a 'how is my position doing' question at get_user_summary. Omit chainId to cover every chain v4 serves, stated under 'chainsCovered'. Zero balances are left out. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_risk_premium` (~195 tokens) How a position's risk premium is built up (v4) v4 only. The per-collateral breakdown of one position's risk premium: how much of the position each token is, and the weight the protocol charges for it. 'currentRiskPremiumWeightPct' is the weight in force now and 'latestRiskPremiumWeightPct' the one the protocol has moved to, so the two differing is the point - the position has been repriced and has not caught up, and the borrow rate will follow. Use it to answer why a v4 borrow costs what it does when the reserve's headline rate says otherwise. 'spokeId' comes from get_user_positions. Input parameters: - `spokeId` (string, required): spokeId from get_user_positions (not the spoke address). - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_spoke_configs` (~207 tokens) What one venue will accept, per asset (v4) v4 only. Per-asset limits for one (hub, spoke) pair: supply and borrow caps, whether the asset is active there, whether it is halted, and the risk premium threshold. Read it before sizing a large position, because a cap is what turns a valid-looking supply into a reverting one. 'halted' stops supply, borrow, withdraw and repay alike; an asset that is simply not active was never configured on that spoke, which is a different thing from one that was turned off. Both ids come from get_reserve_details for any reserve, which is the route that works before a wallet holds anything on that spoke. Input parameters: - `hubId` (string, required): hubId from get_reserve_details or get_hubs. - `spokeId` (string, required): spokeId from get_reserve_details, or from get_user_positions (not the spoke address). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_multichain_asset` (~194 tokens) One asset wherever v4 holds it v4 only. One asset across every v4 chain at once, by symbol: what is supplied and borrowed on each, the rate on each, and a summary naming the best supply rate and the lowest borrow rate anywhere. This is the tool for 'where should I supply USDC' and for 'is this cheaper to borrow somewhere else' - one call rather than reading get_markets per chain. It answers about the asset, not about a wallet. Omit chainId to cover every chain v4 serves, stated under 'chainsCovered'. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `symbol` (string, required): Token symbol, e.g. 'USDC'. Case-insensitive. - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `prepare_liquidation` (~309 tokens) Prepare a liquidation Build an unsigned transaction to liquidate an unhealthy position: repay part of someone else's debt and receive their collateral plus the liquidation bonus. Only viable while that position's health factor is below 1 (check get_user_summary for the target). v4 takes collateral and debt as reserveIds; v3 takes collateralToken and debtToken addresses plus chainId. Non-custodial: the liquidator signs and submits. Input parameters: - `amount` (string): Debt to cover, in main units. Omit only if max=true. - `chainId` (number): v3 only: chain id (positive integer). - `collateral` (string): v4 only: reserveId of the collateral to seize. - `collateralToken` (string): v3 only: collateral token address. - `debt` (string): v4 only: reserveId of the debt to repay. - `debtToken` (string): v3 only: debt token address. - `liquidator` (string): v4 only: the wallet performing the liquidation (0x, 40 hex). On v3 the liquidator is whoever signs the returned transaction. - `market` (string): v3 only: market pool address (from get_markets). - `max` (boolean): Cover the maximum the protocol allows. - `user` (string, required): The wallet being liquidated (0x, 40 hex). - `version` (string, required): Protocol version (v3 or v4; required, no 'all' for actions). ### `prepare_set_emode` (~271 tokens) Prepare an eMode change (v3) v3 only. Build an unsigned transaction to switch a wallet's Aave v3 eMode category. eMode groups correlated assets (ETH-correlated, stablecoins) so they borrow against each other at a higher LTV, raising borrowing power at the cost of restricting which assets the position may hold. Take categoryId from get_emode_categories for that market; pass 0 to turn eMode off. v3 only: v4 replaces eMode with risk premium and dynamic config. Input parameters: - `categoryId` (number, required): eMode categoryId from get_emode_categories, or 0 to disable. - `chainId` (number, required): Chain id (positive integer). - `market` (string, required): v3 only: market pool address, from a get_markets row in this session. It cannot be recalled: an Aave pool address you already recognise belongs to another deployment (v2, or another chain) and is rej… - `sender` (string, required): Sender wallet address (0x, 40 hex): the wallet that will sign, as the user named it in this session. If no wallet has been named, ask for it; never substitute a placeholder, which is rejected. - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `prepare_action` (~959 tokens) Prepare supply / borrow / withdraw / repay Build an unsigned transaction to supply, borrow, withdraw, or repay on Aave (v3 or v4). Returns an execution plan (a transaction to sign, or an approval step first). For a borrow or a withdraw, call preview_action first and read its warnings: the simulation is free, it commits nothing, and it is the cheapest way to learn the position cannot take the action - a withdraw of collateral pinned by an open borrow is the case that catches callers out. Non-custodial: the user's wallet signs and submits; nothing is sent here, so building one is free and commits the user to nothing. Choose sensible values, build it, and say what you chose - do not stop to ask the user first, because they see the transaction before they sign it. Send only the arguments that apply and leave every other one out: v4 takes 'reserveId'; v3 takes 'market' + 'token' + 'chainId'; 'max' is for withdraw and repay; 'enableCollateral' is for supply. Never send 'permitSignature' or 'permitDeadline' on a first call: they exist only to re-issue a call after an approval step handed you a permit to sign under 'bySignature', so if you have not been handed one, leave both out. Two things that catch callers out: a supply does NOT become collateral unless you pass enableCollateral:true, and without collateral the borrowing power is zero so any later borrow is refused; and collateral pinned by an open borrow cannot be withdrawn, so a withdraw has to repay first. After sending any transaction, poll get_transaction_processed before building the dependent next one. Input parameters: - `action` (string, required): Which action to build. - `amount` (string): Amount in MAIN units (e.g. '10.5'), never base units: 100000 base units of a 6-decimal token is '0.1', not '100'. Convert before sending if the user stated base units. Omit only if max=true. - `chainId` (number): v3 only: chain id (positive integer). - `enableCollateral` (boolean): supply only: also enable the asset as collateral (default false). Leave it off and the supply earns yield but backs no borrowing, so a following borrow fails for exceeding a borrowing power of zero.… - `market` (string): v3 only: market pool address, from a get_markets row in this session. It cannot be recalled: an Aave pool address you already recognise belongs to another deployment (v2, or another chain) and is rej… - `max` (boolean): Withdraw/repay only: use the entire balance/debt. - `native` (boolean): Use the chain's native gas token instead of an ERC-20. - `permitDeadline` (number): Unix seconds, the 'deadline' from the permit message that was signed. Required with permitSignature. - `permitSignature` (string): Send this only to replay a call after an approval handed you a permit to sign: the user's signature (0x...) over the PermitTypedData that approval carried under 'bySignature', on either version. It r… - `reserveId` (string): v4 only: the opaque reserveId, copied verbatim from a get_markets row or a get_position_items item in this session (e.g. 'MTo6MHg5NGU3...Ojo1') - it cannot be constructed or recalled, so fetch one be… - `sender` (string, required): Sender wallet address (0x, 40 hex): the wallet that will sign, as the user named it in this session. If no wallet has been named, ask for it; never substitute a placeholder, which is rejected. - `token` (string): v3 only: underlying token address. - `version` (string): Optional: inferred from the reserve selector ('reserveId' is v4, 'market'+'token'+'chainId' is v3). Send it to be explicit, or if you somehow set both. ### `prepare_set_collateral` (~513 tokens) Prepare enable/disable collateral Build an unsigned transaction to change whether a supplied asset is used as collateral (v3 or v4). On v4 'enable' sets the state; on v3 it toggles the current state, and a toggle that would invert a stated 'enable' is refused. The wallet must already hold a supply in that reserve: this sets a flag on it, nothing moves. Disabling can leave a position liquidatable, so it is simulated first: a health factor under 1 after is refused, just above 1 carries a warning to relay, and a build whose simulation could not run says so under 'warnings'. Enabling needs no check. Non-custodial. Input parameters: - `chainId` (number): v3 only: chain id (positive integer). - `enable` (boolean): v4: true to enable, false to disable (default true). v3 toggles the current state; send it on v3 as the intended end state, and a toggle that would produce the opposite is refused. - `market` (string): v3 only: market pool address. - `reserveId` (string): v4 only: the opaque reserveId, copied verbatim from a get_markets row or a get_position_items item in this session (e.g. 'MTo6MHg5NGU3...Ojo1') - it cannot be constructed or recalled, so fetch one be… - `sender` (string, required): Sender wallet address (0x, 40 hex): the wallet that will sign, as the user named it in this session. If no wallet has been named, ask for it; never substitute a placeholder, which is rejected. - `token` (string): v3 only: underlying token address. - `version` (string): Optional: inferred from the reserve selector ('reserveId' is v4, 'market'+'token'+'chainId' is v3). Send it to be explicit, or if you somehow set both. ### `preview_action` (~792 tokens) Preview an action (simulate before executing) Simulate a supply/borrow/withdraw/repay against a wallet's position WITHOUT executing, on v3 or v4. Always do this before a borrow or a withdraw. Send only the arguments that apply: v4 takes 'reserveId'; v3 takes 'market' + 'token' + 'chainId'; 'max' is for withdraw and repay; 'native' works on both; 'enableCollateral' is v4 only. Both versions answer with 'healthFactorBefore' and 'healthFactorAfter'; v4 also returns net APY, risk premium, net collateral, net balance, projected earnings and both borrowing-power figures, each as a matching Before/After pair, plus 'rewardsAcquired' / 'rewardsAbandoned' when the action changes rewards. v3 has the two health factors and nothing else. Either version also returns 'warnings' when the action would not actually succeed - an error level there means the prepare step will refuse it, so fix the inputs rather than building it. Simulate first even when you intend to build immediately: this is the cheapest way to find out that an action cannot succeed, and it commits nothing. It reports the position's own limits and not token allowances, so a clean simulation says the position allows this, not that no approval step remains. Input parameters: - `action` (string, required): Action to simulate. - `amount` (string): Amount in MAIN units (e.g. '10.5'), never base units: 100000 base units of a 6-decimal token is '0.1', not '100'. Convert before sending if the user stated base units. Omit only if max=true. - `chainId` (number): v3 only: chain id (positive integer). - `enableCollateral` (boolean): v4 supply only: also enable as collateral. - `market` (string): v3 only: market pool address, from a get_markets row in this session. It cannot be recalled: an Aave pool address you already recognise belongs to another deployment (v2, or another chain) and is rej… - `max` (boolean): Withdraw/repay only: use the entire balance/debt. - `native` (boolean): Use the chain's native gas token instead of an ERC-20, on either version. Pass it whenever the action is in the native token, or the balance check below reads the wrapped ERC-20 balance and can refus… - `reserveId` (string): v4 only: the opaque reserveId, copied verbatim from a get_markets row or a get_position_items item in this session (e.g. 'MTo6MHg5NGU3...Ojo1') - it cannot be constructed or recalled, so fetch one be… - `sender` (string, required): Sender wallet address (0x, 40 hex): the wallet that will sign, as the user named it in this session. If no wallet has been named, ask for it; never substitute a placeholder, which is rejected. - `token` (string): v3 only: underlying token address. - `version` (string): Optional: inferred from the reserve selector ('reserveId' is v4, 'market'+'token'+'chainId' is v3). Send it to be explicit, or if you somehow set both. ### `get_swappable_tokens` (~205 tokens) Get swappable chains and tokens No version argument: token swaps are protocol-agnostic and run on the v4 backend, so they cover the chains that backend serves rather than every chain Aave has a market on. Call this before get_swap_quote to see where a swap can be quoted: with no arguments it sweeps the chains v4 serves and returns each one with how many tokens are swappable there, which is the cheap answer; pass 'chainIds' to get the actual token list, and to ask about a chain the sweep does not reach. A chain under 'chainsNotCovered' could not be read and is worth asking about again; a chain present with 0 swappable tokens answered, and the answer is that nothing can be swapped there. Per reserve, get_markets also returns 'canSwapFrom'. Input parameters: - `chainIds` (array): Optional: chain ids to list tokens for. Omit the argument entirely to get every chain with a count instead of a token list. ### `get_swap_quote` (~348 tokens) Get a token swap quote No version argument: token swaps are protocol-agnostic and run on the v4 backend, so the chains and tokens they cover are that backend's and not every chain Aave holds a market on - a chain it does not serve is rejected outright rather than quoted badly. Call get_swappable_tokens first if you do not already know the chain is covered, or read 'canSwapFrom' on the reserve from get_markets. Quote a token swap: returns the quote with a quoteId to prepare and sign through prepare_order. Input parameters: - `amount` (string, required): Amount in main units (of the sell token for kind=sell, the buy token for kind=buy). - `buy` (string, required): Token to buy: 'native' for the chain's gas token, or an ERC-20 address (0x...). - `chainId` (number, required): Chain id (positive integer). - `kind` (string): sell = spend an exact sell amount; buy = receive an exact buy amount. Default sell. - `sell` (string, required): Token to sell: 'native' for the chain's gas token, or an ERC-20 address (0x...). - `slippagePct` (string): Optional max slippage as a percent string, e.g. '0.5' for 0.5%. Omit to accept the backend's suggestion, which moves with market conditions - set this to cap it. - `user` (string, required): Wallet address (0x, 40 hex) of the wallet that will sign, as the user named it in this session. If no wallet has been named, ask for it; never substitute a placeholder, which is rejected. ### `prepare_order` (~287 tokens) Prepare an order (typed data or transaction to sign) No version argument: every order runs on the v4 backend. Which optional arguments apply is decided by the quote that minted the quoteId, so decide before emitting any: a quote that offered a 'bySignature' permit takes permitSignature + permitDeadline once that permit is signed; route:'transaction' applies only to a SwapByTransaction... quote (a native-token sell) and takes no signatures at all. Never send a signature argument unsolicited - it exists to carry back a signature the quote explicitly asked for. Non-custodial: this returns what the user signs, and commits nothing. Input parameters: - `permitDeadline` (number): Unix seconds, the 'deadline' from the permit message that was signed. Required with permitSignature. - `permitSignature` (string): The user's signature (0x...) over the PermitTypedData a quote's approvals offered under 'bySignature'. Sending it skips the approval transaction. Never send it on a first call - you only have one onc… - `quoteId` (string, required): quoteId from get_swap_quote. - `route` (string): Which execution route the quote needs, taken from its __typename: 'intent' (default) for everything except SwapByTransaction..., which takes 'transaction' and returns the on-chain transaction plus an… ### `submit_signed_order` (~91 tokens) Submit a user-signed order No version argument: every order runs on the v4 backend. Relay an order the user already signed (quoteId + signature from prepare_order's PreparedOrder). State-changing: this posts the order. The server never signs. Input parameters: - `quoteId` (string, required): quoteId (the newQuoteId returned by prepare_order). - `signature` (string, required): The user's EIP-712 signature (0x...). ### `prepare_cancel_order` (~122 tokens) Prepare an order cancellation (EIP-712 to sign) No version argument. From an orderId, return the EIP-712 cancellation for the user to sign; cancel_order relays the result. This is the gasless route for an intent order; an order that went on chain (a native-token sell) is cancelled with cancel_order's transaction route instead, which needs no signature and no prepare. Non-custodial: this returns what the user signs, and commits nothing. Input parameters: - `orderId` (string, required): The order id (from submit_signed_order's receipt, get_order_status or get_pending_orders). ### `cancel_order` (~158 tokens) Cancel an order No version argument. Cancel an open order. Two routes: pass 'signature' (signed over what prepare_cancel_order returned) to relay a gasless cancellation - state-changing, and the route an intent order takes; omit 'signature' entirely for the on-chain route, which returns the cancel transaction for the user to sign and send, the route an order that went on chain (a native-token sell) takes. The server never signs. Input parameters: - `orderId` (string, required): The order id (from submit_signed_order's receipt, get_order_status or get_pending_orders). - `signature` (string): The user's cancellation signature (0x...) over prepare_cancel_order's typed data. Omit the argument entirely to get the on-chain cancel transaction instead. ### `get_order_status` (~85 tokens) Get order status No version argument. Status of an order by id: open, pending signature, fulfilled, cancelled or expired, with the operation it performs and an explorer link. 'order' comes back null when the API cannot see that id, which a wrong id and a just-submitted order both produce - poll again briefly before concluding. Input parameters: - `orderId` (string, required): The order id. ### `get_pending_orders` (~198 tokens) List a wallet's orders No version argument: orders run on the v4 backend, so this covers the chains that backend serves and names them under 'chainsCovered'; a chain under 'chainsNotCovered' could not be read and is worth asking about again. A wallet's orders, newest first. Despite the name it returns every status unless 'statuses' narrows it - pass ['open','pending_signature'] for what is still actionable. Each chain returns its newest 50 orders; 'more: true' on a chain means older history exists beyond what was returned. Use this to recover an orderId you no longer have. Input parameters: - `chainIds` (array): Optional: chain ids to read. Omit the argument entirely to sweep every chain the backend serves. - `statuses` (array): Optional: statuses to keep. Omit the argument entirely for all of them. - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). ### `get_user_rewards` (~218 tokens) Get user rewards A wallet's claimable Aave rewards on v3, v4, or both. Covers incentive programmes distributed through Merit as well as the protocol's own, so a GHO or sGHO incentive shows up here rather than anywhere separate. Omit chainId and both versions cover every chain they serve, stating which under 'chainsCovered' and tagging each row with its own chainId. 'chainsNotServed', when present, lists chains this API holds no market on: Aave supports the chain, there is nothing here to read, so an empty result for one of them is not an answer about that chain, and asking again returns the same nothing. On v3 the response also carries the transaction that claims them. Input parameters: - `chainId` (number): Chain id to restrict to, e.g. 1 for Ethereum. Omit the argument entirely to cover every chain. - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Protocol version (default all). ### `prepare_claim_rewards` (~166 tokens) Prepare a rewards claim (V4) v4 only. Build an unsigned transaction to claim a wallet's V4 rewards. If 'ids' is omitted, claims everything currently claimable. Non-custodial. There is no v3 equivalent because a v3 claim needs no second call: get_user_rewards already returns the transaction. Input parameters: - `chainId` (number): Chain id to claim on. Required unless 'ids' is omitted, in which case the chain is discovered from the wallet's claimable rewards. - `ids` (array): Optional reward ids (from get_user_rewards). Omit to claim all claimable. - `user` (string, required): Wallet address (0x, 40 hex). - `version` (string): Optional, and only 'v4': this tool exists on v4 only. ### `get_sgho_vault` (~317 tokens) Get the sGHO savings vault v3 only, Ethereum only. Savings GHO (sGHO) is an ERC-4626 vault that pays a target rate on deposited GHO. Returns the vault's targetRatePct, totalAssets, supplyCap and paused flag; pass 'user' to add that wallet's shares, their GHO value, its GHO wallet balance, and the maxDeposit / maxWithdraw the contract allows right now. Quote what a deposit earns from 'combinedAprPct' when it is present: a Merit campaign sometimes pays 'incentiveAprPct' on top of the target rate, and the two differ in how they arrive - the target rate accrues inside the vault and needs no action, the incentive is claimed separately through the link in 'incentive' and can end at any time. All three are absent together when no campaign is running, which is the usual state, and then targetRatePct is the whole rate. Two things this rate is not: it is set by governance rather than by utilisation, so it does not move with borrowing demand, and it is not a lending position - sGHO earns yield but is not Aave collateral and cannot be borrowed against. Deposits and withdrawals are instant, with no cooldown. Input parameters: - `user` (string): Optional wallet address (0x, 40 hex) to include its position. Omit the argument entirely for vault state only. - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `get_sgho_preview` (~230 tokens) Preview an sGHO deposit or withdrawal v3 only, Ethereum only. Convert between GHO and sGHO shares at the vault's current index, without building anything. Which unit 'amount' is in depends on the action: 'deposit' takes GHO and returns the shares it would mint; 'withdraw' takes sGHO SHARES and returns the GHO they would return. Call this with action 'withdraw' before prepare_sgho_action, because that withdrawal is denominated in shares and not in GHO - passing a GHO figure there withdraws the wrong amount. When quoting what a user-named amount converts to, use this preview rather than multiplying by the rate: the index moves. Input parameters: - `action` (string, required): 'deposit' converts GHO to shares; 'withdraw' converts shares to GHO. - `amount` (string, required): Amount in main units (e.g. '10.5'): GHO when action is 'deposit', sGHO shares when it is 'withdraw'. - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `prepare_sgho_action` (~335 tokens) Prepare an sGHO deposit or withdrawal v3 only, Ethereum only. Build an unsigned transaction to deposit GHO into the sGHO vault or withdraw from it. Non-custodial: the user's wallet signs and submits, so building one commits them to nothing. Choose sensible values, build it, and say what you chose rather than stopping to ask. Send only the arguments that apply and leave every other one out. On 'deposit', 'amount' is GHO. On 'withdraw', 'amount' is a number of sGHO SHARES, not GHO - get_sgho_preview converts - or pass max:true to redeem the whole position and omit 'amount'. 'recipient' is only for sending the result to a different address than the sender. A deposit needs a GHO allowance: when the result is ApprovalRequired, submit its 'approval' transaction and then the 'originalTransaction' beside it, in that order. Input parameters: - `action` (string, required): Which side to build: deposit GHO, or withdraw by redeeming shares. - `amount` (string): Amount in main units: GHO to deposit, or sGHO shares to redeem. Omit only when max=true. - `max` (boolean): withdraw only: redeem the entire share balance. - `recipient` (string): Optional address to receive the shares (deposit) or the GHO (withdraw). Omit the argument entirely to use the sender. - `sender` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `prepare_stkgho_migrate` (~152 tokens) Prepare a stkGHO migration into sGHO v3 only, Ethereum only. Build an unsigned transaction that moves a wallet's stkGHO out of the Safety Module and into the sGHO vault. All-or-nothing: it covers the entire stkGHO position and takes no amount. Call get_safety_module with the same wallet first and quote the legacy stkGHO balance it returns, so the user knows what they are moving before they sign. Non-custodial, so build it rather than asking permission first. Input parameters: - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `get_safety_module` (~323 tokens) Get the Safety Module Aave's Safety Module, Ethereum only, and the one product here that is read from the contracts rather than the Aave API. Returns both generations: 'umbrella', the per-asset stake tokens securing the protocol now, each with its total staked, target liquidity, reward APY, cooldown and unstake window; and 'legacy', the stkAAVE / stkGHO / stkABPT tokens alongside it. Pass 'user' to add that wallet's staked balance, claimable rewards and cooldown state on both. Staking is not supplying: a stake earns its own rewards, is not Aave collateral, and cannot be borrowed against - but it is part of a wallet's Aave position, so a portfolio answer that omits it is wrong. USD figures are Aave's oracle price for the staked asset. Reward APY is what the contracts pay right now, not the configured maximum, because Umbrella scales emission by how far total staked sits from target liquidity. One field decides whether there is an answer here at all: a result carrying 'read': false and no 'umbrella' or 'legacy' means this deployment cannot reach the Safety Module, which is a gap in its configuration and not a fact about the wallet - say the staking part is unavailable rather than reporting the wallet as unstaked, and do not call again. Input parameters: - `user` (string): Optional wallet address (0x, 40 hex) to include its stake, rewards and cooldown. Omit the argument entirely for protocol-wide state only. ### `get_vaults` (~381 tokens) Get Aave Vaults, or a wallet's shares in them v3 only. Name exactly one of 'user', 'owner' or 'vault', and read that sentence before choosing arguments: 'user' lists every vault that wallet holds shares in and is the portfolio question, 'owner' lists what one curator runs, 'vault' reads a single vault by address and is the only one that also needs 'chainId'. An Aave Vault is an ERC-4626 vault a curator deploys over one v3 reserve, taking a fee on the yield, so its APR is below that reserve's by the fee. Shares in one are an Aave position that no market read returns. There is no catalogue to browse: this API has no "list every vault" call, so a request to find the best vault cannot be answered here - you can only look up vaults whose address, owner or holder you already have. 'aprPct' is already net of 'feePct'. Input parameters: - `chainId` (number): Chain the vault is on, e.g. 1 for Ethereum. - `cursor` (string): 'nextCursor' from a previous call, to continue a 'user' or 'owner' listing. Omit the argument entirely for the first page. - `owner` (string): Curator address (0x, 40 hex) to list the vaults it owns. Omit the argument entirely when using another selector. - `user` (string): Wallet address (0x, 40 hex) to list its vault positions. Omit the argument entirely when using another selector. - `vault` (string): Vault contract address (0x, 40 hex) to read one vault. Needs 'chainId'. Omit the argument entirely when using another selector. - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `get_vault_preview` (~254 tokens) Preview an Aave Vault deposit, mint, withdraw or redeem v3 only. Convert between a vault's assets and its shares at the current index, without building anything. Which unit 'amount' is in depends on the action, and this is the thing to get right: 'deposit' takes ASSETS and returns the shares minted, 'mint' takes SHARES and returns the assets it would cost, 'withdraw' takes ASSETS and returns the shares burnt, 'redeem' takes SHARES and returns the assets returned. Use the assets pair to fix the token amount exactly and the shares pair to fix the share count exactly. Call this before prepare_vault_action rather than multiplying by a rate: the index moves. Input parameters: - `action` (string, required): Which conversion: deposit/withdraw speak assets, mint/redeem speak shares. - `amount` (string, required): Amount in main units (e.g. '10.5'), in the unit the action takes. - `chainId` (number, required): Chain the vault is on, e.g. 1 for Ethereum. - `vault` (string, required): Vault contract address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `prepare_vault_action` (~370 tokens) Prepare an Aave Vault deposit, mint, withdraw or redeem v3 only. Build an unsigned transaction to move value in or out of an Aave Vault. Non-custodial: the user's wallet signs and submits, so building one commits them to nothing - choose sensible values, build it, and say what you chose rather than stopping to ask. Which unit 'amount' is in depends on the action: 'deposit' and 'withdraw' take ASSETS, 'mint' and 'redeem' take SHARES. Set 'asAToken' true to move the reserve's aToken instead of its underlying token, which skips supplying first when the wallet already holds the aToken. 'recipient' is only for sending the result to a different address than the sender. A deposit or mint needs an allowance: when the result is ApprovalRequired, submit its 'approval' transaction, wait for it to be mined, then submit the 'originalTransaction' beside it. Input parameters: - `action` (string, required): Which side to build: deposit/withdraw in assets, mint/redeem in shares. - `amount` (string, required): Amount in main units, in the unit the action takes (assets or shares). - `asAToken` (boolean): Move the reserve's aToken rather than its underlying token (default false). - `chainId` (number, required): Chain the vault is on, e.g. 1 for Ethereum. - `recipient` (string): Optional address to receive the shares or the assets. Omit the argument entirely to use the sender. - `sender` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `vault` (string, required): Vault contract address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v3': this tool exists on v3 only. ### `get_vault_activity` (~220 tokens) A wallet's history in one Aave Vault v3 only. How one wallet's position in one vault moved over a window: total yield earned, plus a per-sample breakdown of balance, deposits and withdrawals. Use it to answer 'what has this vault actually paid me', which the current APR cannot - the APR is the rate now, this is what was received. Needs the vault address and its chain, both of which get_vaults returns. Input parameters: - `chainId` (number, required): Chain the vault is on, e.g. 1 for Ethereum. - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). - `vault` (string, required): Vault contract address, 0x-prefixed (40 hex chars). - `version` (string): Optional, and only 'v3': this tool exists on v3 only. - `window` (string): How far back to go (default week). These are the only buckets this endpoint has: there is no day and no six months, so ask for one of these rather than expecting the nearest. ### `search_governance_proposals` (~261 tokens) Search Aave DAO proposals Takes no 'version' argument (DAO governance is not per-market; Governance V3 is unrelated to Aave v3/v4 markets). List or search Aave DAO governance proposals. Filter by 'state' to answer "what is live right now" (state 'active') or "what is waiting to execute" (state 'queued'). Pass 'search' for full-text over title, author and description, ranked by relevance. Vote tallies are in AAVE. Input parameters: - `includeSummaries` (boolean): Include each proposal's ~500-char summary (default false). Adds roughly 500 bytes per row, so ask for it when summarising a list rather than when looking one up. - `limit` (number): Max rows to return (1-100, default 20). - `offset` (number): Rows to skip, for paging (default 0). - `search` (string): Optional full-text query, e.g. 'GHO' or 'Chaos Labs'. Combines with 'state'. - `state` (string): Filter to one lifecycle state. 'active' is open for voting, 'queued' passed and waits out the timelock, 'executed' is done. Omit for every state. ### `get_governance_proposal` (~183 tokens) Get an Aave DAO proposal Takes no 'version' argument (DAO governance is not per-market; Governance V3 is unrelated to Aave v3/v4 markets). Full detail for one Aave DAO governance proposal: state, author, vote tallies, quorum, timings and the transaction hash of each lifecycle step. 'quorumMet' and 'differentialMet' are the two conditions a proposal must meet to pass. Vote tallies and quorum are both in AAVE, so they compare directly. The long IPFS write-up is omitted unless includeDescription is true. Input parameters: - `includeDescription` (boolean): Include the full proposal markdown from IPFS (default false). It runs to many KB; the short description is always returned. - `proposalId` (string, required): Proposal id as a decimal string, e.g. '508'. From search_governance_proposals. ### `get_proposal_votes` (~181 tokens) Get votes on a proposal Takes no 'version' argument (DAO governance is not per-market; Governance V3 is unrelated to Aave v3/v4 markets). Who voted on an Aave DAO proposal and with how much power, largest voter first. Pass support=true or false to list only one side. 'totals' covers every vote on the proposal, not just the page returned. Voting power is in AAVE. Input parameters: - `limit` (number): Max rows to return (1-100, default 20). - `offset` (number): Rows to skip, for paging (default 0). - `proposalId` (string, required): Proposal id as a decimal string, e.g. '508'. From search_governance_proposals. - `support` (boolean): Filter to one side: true for votes in favour, false for against. Omit for both. ### `get_user_vote` (~127 tokens) Get a wallet's vote on a proposal Takes no 'version' argument (DAO governance is not per-market; Governance V3 is unrelated to Aave v3/v4 markets). How one wallet voted on one Aave DAO proposal, and with how much power. Returns voted:false when that wallet did not vote, which is an answer rather than an error. Voting power is in AAVE. Input parameters: - `proposalId` (string, required): Proposal id as a decimal string, e.g. '508'. From search_governance_proposals. - `user` (string, required): Wallet address, 0x-prefixed (40 hex chars). ### `get_proposal_payloads` (~138 tokens) Get a proposal's payloads Takes no 'version' argument (DAO governance is not per-market; Governance V3 is unrelated to Aave v3/v4 markets). What an Aave DAO proposal actually executes: one payload per target chain, each with its execution state, timelock delay, grace period and the contract calls it makes. Use this to tell whether a passed proposal has landed on every chain yet: a proposal can be 'executed' on mainnet while a payload is still 'queued' elsewhere. Input parameters: - `proposalId` (string, required): Proposal id as a decimal string, e.g. '508'. From search_governance_proposals. ### `get_aave_guide` (~253 tokens) Aave protocol & usage guide Applies to both versions. Protocol and usage guidance for building Aave flows correctly. Read the relevant topic before constructing an action. Topics: overview, v4, v3, positions, health-factor, risks, ids, amounts, prices, signing, swaps, gho, safety-module, portfolio, governance, rewards, tools, docs. Also the place to come when no tool seems to fit what you were asked for: send 'lookingFor' and this answers with the full capability list rather than leaving you to guess from tool names. For deep protocol docs, point the user to https://docs.aave.com. Input parameters: - `lookingFor` (string): Send this when you looked for a tool to do something and did not find one, or when a flow dead-ended: describe what you were trying to do, in a few words. It returns the full capability list so you c… - `topic` (string): Guide topic (default overview). Read 'v3' or 'v4' before building an action on that version. ### `get_started` (~59 tokens) What this connector can do Applies to both versions. What this Aave server can do: the tool families, the ready-made prompts, and the guide topics. Call this when an Aave conversation opens with no specific request and answer from it, rather than guessing from tool names. ## Diagnostics Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/com-aave-mcp/mcp#diagnostics ## Score history - 2026-09-23: 79 - 2026-09-22: 78 - 2026-09-21: 78 - 2026-09-20: 77 - 2026-09-19: 77 - 2026-09-18: 76 - 2026-09-17: 76 - 2026-09-16: 75 - 2026-09-15: 75 - 2026-09-14: 74 - 2026-09-13: 74 - 2026-09-12: 73 - 2026-09-11: 73 - 2026-09-10: 73 - 2026-09-09: 72 - 2026-09-08: 72 - 2026-09-07: 71 - 2026-09-06: 71 - 2026-09-05: 70 - 2026-09-04: 70 - 2026-09-03: 70 - 2026-09-02: 69 ## Common questions ### What is the Aave MCP server? Aave MCP is listed in the public MCP registry as com.aave/mcp. Official Aave MCP for V3 and V4 markets, positions, governance, and transaction preparation. This page covers its hosted endpoint (https://mcp.aave.com). ### Is the Aave MCP server safe to use? Aave MCP scores 79 out of 100 on VerifyMCP. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm. ### What tools does the Aave MCP server expose? Aave MCP exposes 53 tools: get_chains, get_markets, get_emode_categories, get_user_positions, get_position_items, and 48 more. Their descriptions and schemas cost roughly 13,676 tokens of context every time the server is loaded. ### Does the Aave MCP server require authentication? No. We connected to Aave MCP without credentials and it answered, so anything it exposes is reachable by anyone who knows the address. ### Is the Aave MCP server still maintained? Aave MCP is still listed as active in the MCP registry. We last reached this channel on 23 September 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced. ## Links - Remote endpoint: https://mcp.aave.com/ - Website: https://aave.com/ - Changelog RSS feed: https://verifymcp.io/servers/com-aave-mcp/mcp.xml - Changelog JSON feed: https://verifymcp.io/servers/com-aave-mcp/mcp.json - HTML version of this page: https://verifymcp.io/servers/com-aave-mcp/mcp