Every pump.fun curve trade since 09-11: tape, launches, forensics, devs, odds + Meteora DLMM pools.
Scored across 12 tools
Each tool has a clearly distinct purpose: pump_pulse (live market), pump_history (trends), pump_token (token forensics), pump_dev (creator history), pump_wallet (wallet PnL), pump_odds (graduation probability), copy_backtest (copy simulation), smart_money (wallet leaderboard), pump_launches (new mints), and DLMM tools (dlmm_pool, dlmm_positions, dlmm_whales). Descriptions explicitly differentiate and cross-reference to avoid confusion.
Most tools follow a clear prefix pattern (pump_* for pump.fun, dlmm_* for Meteora DLMM) with descriptive suffixes. The only slight inconsistency is copy_backtest and smart_money not sharing a prefix, but they are still readable and contextually clear.
12 tools are well-scoped for covering two domains (pump.fun analytics and Meteora DLMM) with distinct analytical perspectives. Each tool earns its place without redundancy.
The surface covers comprehensive analytics for pump.fun (market, token, creator, wallet, copy simulation, leaderboard, odds) and DLMM (pool, positions, whales). Some gaps exist, such as direct DLMM pool discovery or creation tools, but these are outside the analytical focus. Cross-server references indicate additional layers exist elsewhere.
Would copying this wallet have made money? Replays its pump.fun entries and exits 0–25 slots late on the recorded curve and shows where the edge dies. Takes the wallet's pump.fun positions over the last 1–14 days and simulates a copier that buys the same token and sells when the leader sells, arriving 0, 1, 2, 5, 10 or 25 slots (0–10 s) late. Fills use the bonding-curve reserves our tape recorded at that moment, including curve fees and the copy's own price impact. Returns PnL, win rate and ROI per latency, the leader's own median ROI for contrast, and a verdict: copyable, latency-sensitive (edge gone by N slots), or not copyable. A wallet that looks great on its own PnL often loses money for anyone who follows it. When to use: After pump_wallet or smart_money, to check copying survives realistic delay. Price: $0.03 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| days | No | How many days of the wallet's history to replay. Range 1-14. Default 3. | |
| exit | No | sell on the leader's first sell, or once it has sold 90%. Default "first_sell". | first_sell |
| wallet | Yes | leader wallet to copy. Base58 Solana address, 32-44 chars. | |
| size_sol | No | SOL per copied entry. Range 0.01-10. Default 0.1. | |
| max_positions | No | most recent positions to replay. Range 10-300. Default 150. |
| Name | Required | Description |
|---|---|---|
| note | No | |
| found | No | |
| label | No | |
| setup | No | |
| by_lag | No | |
| wallet | No | |
| verdict | No | |
| fast_flips_pct | No | |
| biggest_at_2_slots | No | |
| leader_median_roi_pct | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Goes well beyond the annotations (readOnly/idempotent/non-destructive): it discloses the fill model (bonding-curve reserves from the recorded tape, including curve fees and the copy's own price impact), the latency grid, the returned verdict categories, and pricing ($0.03/call, 10 free/day, x402 payment-required result). It also clarifies that upstream errors return isError and are not charged.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loads the core question and the mechanism, then layers simulation details, output, verdict, and operational notes (cost, errors). Dense but mostly non-redundant; the motivating line about wallets that look great on their own PnL earns its place as rationale, though the prose runs long.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given an output schema exists, the description need not explain returns, yet it summarizes PnL/win-rate/ROI per latency and the verdict anyway. Combined with cost, error behavior, and the when-to-use hook, nothing an agent needs to call it correctly is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so every parameter is already documented with ranges, defaults and the exit enum. The description reinforces the meaning ('last 1-14 days', 'sells when the leader sells') but adds no syntax or semantics beyond what the schema provides. Baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb+resource (replays a wallet's pump.fun entries/exits) and frames it with a concrete question ('Would copying this wallet have made money?'). It explicitly distinguishes itself from sibling tools by naming pump_wallet and smart_money as the upstream steps, so an agent can place it in the workflow without opening the schema.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
'When to use: After pump_wallet or smart_money, to check copying survives realistic delay' gives both the trigger condition and the alternatives it complements. It also states cost and error semantics, leaving no routing or invocation ambiguity.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
State of a Meteora DLMM pool: price, bin step, fees, TVL, volume and fees by window (30 m–24 h), 24 h fee yield and its annualized APR/APY. For LP agents choosing where to provide liquidity. When to use: For one pool's state; for a wallet's positions use dlmm_positions. Price: $0.002 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| pool | Yes | DLMM pool (pair) address. Base58 Solana address, 32-44 chars. |
| Name | Required | Description |
|---|---|---|
| fees | No | |
| name | No | |
| pool | No | |
| config | No | |
| volume | No | |
| apr_pct | No | |
| token_x | No | |
| token_y | No | |
| tvl_usd | No | |
| reserves | No | |
| created_at | No | |
| yield_note | No | |
| current_price | No | |
| fee_tvl_ratio | No | |
| dynamic_fee_pct | No | |
| fee_yield_24h_pct | No | |
| apy_pct_daily_compounding | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Beyond the annotations (which already declare read-only/idempotent/open-world), the description discloses cost and quota behavior ($0.002 per call, 10 free/day, then an x402 payment-required result), and error semantics (isError with a message on invalid input or upstream failure, and that failures are not charged). These are operational traits an agent cannot obtain from the annotations or schema.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loads what the tool returns, then usage, then pricing, then error behavior — a sensible priority order with no filler sentences. It is a dense single paragraph, but every clause carries information.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
An output schema exists, so return-value explanation is unnecessary, and the description still covers scope, sibling routing, pricing/quota, and error handling for a one-parameter tool. Nothing an agent needs to call this correctly is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100% and the single 'pool' parameter is fully documented in the schema (base58 Solana address, 32-44 chars) plus a regex pattern. The description adds no format or syntax detail beyond that, so the schema does the heavy lifting — baseline 3 is correct.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific resource (Meteora DLMM pool state) and enumerates exactly which state fields are returned (price, bin step, fees, TVL, volume, fee yield, APR/APY), so an agent knows precisely what it gets. It also explicitly separates itself from the sibling dlmm_positions by scope (one pool vs a wallet's positions).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Gives an explicit target audience ('For LP agents choosing where to provide liquidity') plus a when-to-use rule: 'For one pool's state; for a wallet's positions use dlmm_positions.' The alternative tool and the condition selecting it are named directly, leaving nothing to inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Open Meteora DLMM positions of any wallet: value, unclaimed fees, PnL, in/out of range — or its closed-position history. Set history=true for closed positions with deposits, withdrawals, fees and PnL per pool. When to use: For a wallet's Meteora positions; to find skilled LPs use dlmm_whales. Price: $0.005 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| page | No | Page of closed-position history (only with history=true). Range 1-50. Default 1. | |
| wallet | Yes | wallet address. Base58 Solana address, 32-44 chars. | |
| history | No | true = closed-position history instead of open positions. Default false. |
| Name | Required | Description |
|---|---|---|
| mode | No | |
| page | No | |
| pools | No | |
| total | No | |
| wallet | No | |
| has_next | No | |
| sol_price | No | |
| pools_total | No | |
| positions_total | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare read-only, idempotent, open-world and non-destructive. The description adds substantial context beyond them: per-call pricing, a 10-free/day quota, an x402 payment-required fallback, and error semantics (isError on invalid input or upstream failure, and that failures are not charged). Gaps remain only around auth details.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loads the core capability before cost and error notes, and every sentence carries information. It is dense, packing pricing, quota, routing and error handling into one paragraph, but nothing is filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
An output schema exists, so return values need no explanation. The description covers the use case, the closed-position mode, cost/payment path and error behavior, which is everything an agent needs to call it correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so the baseline is 3, but the description meaningfully explains the history flag (closed positions with deposits, withdrawals, fees and PnL per pool), which adds semantic value beyond the schema's terse 'true = closed-position history'.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific resource and scope: open Meteora DLMM positions for any wallet, enumerating returned data (value, unclaimed fees, PnL, in/out of range) and the closed-position variant. It explicitly distinguishes itself from the sibling dlmm_whales, so an agent can route correctly without opening the schema.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The 'When to use' clause names the exact scenario (a wallet's Meteora positions) and the alternative for the adjacent goal (find skilled LPs via dlmm_whales). It also states the condition (history=true) that selects the closed-position mode.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Meteora DLMM liquidity providers ranked by measured on-chain edge: net USD, median multiple, win rate, hold time, size, recency. Produced by our own discovery engine, which decodes DLMM liquidity instructions across high-volume and high-yield pools every few hours and scores each wallet's completed round trips. These are the wallets our live copier watches. Price: $0.02 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | How many ranked LP wallets to return. Range 1-25. Default 10. |
| Name | Required | Description |
|---|---|---|
| method | No | |
| whales | No | |
| generated_at | No | |
| pools_scanned | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover the read-only/idempotent safety profile, and the description adds substantial context beyond them: data provenance (own discovery engine decoding DLMM instructions across high-volume/high-yield pools), refresh cadence (every few hours), cost model ($0.02/call, 10 free/day, x402 payment-required result), and error behavior (isError, not charged). This is exactly the extra behavioral context the bar rewards.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loads the identity and ranking criteria, then provenance, then pricing, then error handling. Dense but every sentence carries distinct information; slightly packed for a single paragraph but no wasted filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
An output schema exists, so return-value explanation is unnecessary, and the description fills the remaining gaps an agent needs: cost/quota, failure semantics, data freshness, and the source of the ranking. Nothing required for correct invocation is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100% for the single 'limit' parameter, including default and 1-25 range, so the schema already carries full semantics. The description adds nothing about the parameter, which is the correct baseline when the schema does the heavy lifting.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb+resource (Meteora DLMM liquidity providers) and spells out the ranking dimensions (net USD, median multiple, win rate, hold time, size, recency). An agent can distinguish it from dlmm_pool, dlmm_positions, and smart_money purely from the description.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Provides useful context ('these are the wallets our live copier watches') that implies a copy-trading/discovery use case, and pricing guidance is explicit. However, it never states when to pick this over siblings like smart_money or dlmm_positions, nor any exclusions. Usage is implied rather than directed.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
A creator wallet's launch history: launches, graduation rate, launch cadence, how fast the dev sells, verdict. Looks up every token the wallet created since recording began and replays the dev's own trades on its 15 most recent launches. When to use: For the creator of a token; pass the dev wallet, not the mint (pump_token reports the dev of a mint). Price: $0.01 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| creator | Yes | creator (dev) wallet. Base58 Solana address, 32-44 chars. |
| Name | Required | Description |
|---|---|---|
| found | No | |
| creator | No | |
| verdict | No | |
| baseline | No | |
| launches | No | |
| graduated | No | |
| last_launch | No | |
| first_launch | No | |
| recent_launches | No | |
| launches_last_24h | No | |
| graduation_rate_pct | No | |
| median_gap_between_launches_s | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare read-only, idempotent, open-world, non-destructive, but the description adds behavior they cannot express: cost ($0.01 per call, 10 free/day, x402 payment-required result), error semantics (isError for invalid input or upstream failure, not charged), and the replay scope (15 most recent launches). That is meaningful disclosure beyond the structured fields.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
It is dense but front-loaded: the returned metrics come first, then usage, then cost, then error behavior. Every sentence carries information, though the run-on enumeration of metrics and the pricing/error clause make it longer than strictly necessary for a single-parameter tool.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given an output schema exists, the description correctly focuses on routing, cost, and failure modes rather than return shape, and the annotations already carry the safety profile. Nothing an agent needs to invoke it or handle payment/error outcomes is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100% with a single well-documented 'creator' property (Base58 pattern, 32-44 chars), so the baseline is 3. The description goes slightly beyond the schema by stressing that this must be the dev wallet rather than the mint address, which disambiguates a common misuse.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description names a specific resource (a creator wallet's launch history) and enumerates exactly what it computes: launches, graduation rate, launch cadence, dev sell speed, and a verdict. It also explicitly distinguishes itself from the closest sibling by stating 'pass the dev wallet, not the mint (pump_token reports the dev of a mint)', so an agent can route correctly without opening either schema.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It gives an explicit 'When to use' clause scoped to the creator of a token, plus a negative instruction ('not the mint') and names the alternative tool that reports the dev of a mint (pump_token). This is precise when/when-not/alternative routing.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
pump.fun over time: launches, graduations, graduation rate, trades, SOL volume and unique wallets per day or hour since 2026-09-11 — with a compare window. The same tape as pump_pulse, sliced into UTC days or hours so an agent can compare the market over time instead of reading one snapshot. Pass compare_from/compare_to to get a second window and the per-period change (e.g. this week vs last week). Closed periods are precomputed from our own live recording of every bonding-curve trade; the open period is marked partial. Also reports the share of trades made by pump.fun's own mayhem agent, so it can be separated from real demand. When to use: For trends over days or hours; for the last 15 minutes use pump_pulse. Price: free. Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| to | No | UTC end, inclusive (default: now) | |
| from | No | UTC start, e.g. 2026-09-18 or 2026-09-18T06 (default: 14 days / 48 hours back) | |
| compare_to | No | end of the comparison window, inclusive | |
| granularity | No | Bucket size: "day" or "hour". Default "day". | day |
| compare_from | No | start of a second window to compare against |
| Name | Required | Description |
|---|---|---|
| rows | No | |
| compare | No | |
| summary | No | |
| coverage | No | |
| timezone | No | |
| definitions | No | |
| granularity | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly, idempotent, openWorld and non-destructive, so safety is covered. The description adds real behavioral context beyond them: closed periods are precomputed from the tool's own recording of bonding-curve trades, the open period is flagged partial, and a share of trades attributable to pump.fun's own mayhem agent is reported separately. It stops short of describing pagination/limits on the number of returned buckets.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loaded with what it returns, then differentiation, compare semantics, provenance, usage and pricing in a logical order. Slightly long and a touch repetitive in restating the pump_pulse relationship, but every sentence carries information an agent needs.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With an output schema present, return values need no explanation, and defaults for from (14 days / 48 hours back) and to (now) are called out explicitly. Provenance, partial-period marking, comparison behavior, cost and error handling together make the definition complete for a 5-parameter, zero-required tool.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the baseline would be 3. The description adds meaning beyond the schema by framing compare_from/compare_to as producing a second window plus a per-period change (e.g. this week vs last week), which the parameter descriptions alone do not convey.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
Names the specific resource (pump.fun metrics: launches, graduations, graduation rate, trades, SOL volume, unique wallets) and the slice it returns (per UTC day or hour) with an explicit compare window. It actively distinguishes itself from the sibling pump_pulse by describing itself as 'the same tape as pump_pulse, sliced into UTC days or hours' rather than a 15-minute snapshot.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Contains an explicit 'When to use' clause: trends over days or hours here, last 15 minutes in pump_pulse. It also states cost ('Price: free') and error/charging behavior, leaving nothing about selection or prerequisites to inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
The newest pump.fun tokens with dev buy, early buyers/volume, curve progress, market cap and creator launch count. Returns up to 50 of the most recent launches (seconds old) with the first-N-seconds flow already aggregated, so an agent can shortlist without replaying the chain. When to use: To discover new mints; then use launch_scan (on the MAD Synapse · Wallets & Risk server, https://agent.maddegen.art/hub/mcp/risk) for a verdict on one of them. Price: $0.002 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | How many of the newest launches to return. Range 1-50. Default 20. | |
| window_s | No | early-flow window after creation. Range 30-3600. Default 300. |
| Name | Required | Description |
|---|---|---|
| launches | No | |
| window_s | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover the safety profile (readOnly, idempotent, non-destructive), and the description adds substantial context beyond them: the up-to-50-row cap, that flow is pre-aggregated for the first N seconds, per-call pricing with a free tier and x402 payment-required behavior, and error semantics (isError, invalid input or upstream failure not charged). That is exactly the extra behavioral detail annotations cannot express.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loads purpose, then usage, then pricing and error behavior, with every sentence carrying information. It is dense and slightly long, but no sentence is filler; the pricing/error block is justified for a paid tool.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With an output schema present, the description need not describe return values, and it still covers freshness ('seconds old'), row cap, pricing, payment flow, error handling, and the recommended next tool. Nothing material for correct invocation is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema already documents limit and window_s fully; baseline would be 3. The description lifts this slightly by explaining the semantics of the window ('the first-N-seconds flow already aggregated' / 'early-flow window after creation'), giving meaning to window_s beyond its range constraints, though it adds nothing for limit.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb and resource (newest pump.fun token launches) and enumerates the payload dimensions (dev buy, early buyers/volume, curve progress, market cap, creator launch count). This clearly separates it from siblings like pump_token (single token) and pump_history, so an agent can pick it without opening the schema.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicitly states 'When to use: To discover new mints' and routes the agent to a follow-up tool (launch_scan on the MAD Synapse server) with its URL. It lacks an explicit when-not or a direct comparison to adjacent siblings such as pump_pulse, so it stops just short of full alternative coverage.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Probability a pump.fun launch graduates, from how comparable launches in our tape actually ended — with the lift over the ~3% base rate. Buckets the launch's first 60 seconds (unique buyers, buy SOL, dev buy) and returns the empirical graduation rate of thousands of comparable recent launches whose outcome is known, shrunk toward coarser buckets when samples are thin. Rebuilt every 6 hours. When to use: For graduation probability only; launch_scan (on the MAD Synapse · Wallets & Risk server, https://agent.maddegen.art/hub/mcp/risk) includes it alongside the other checks. Price: $0.01 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| mint | Yes | pump.fun token mint. Base58 Solana address, 32-44 chars. |
| Name | Required | Description |
|---|---|---|
| mint | No | |
| age_s | No | |
| found | No | |
| model | No | |
| method | No | |
| status | No | |
| symbol | No | |
| swap_url | No | |
| created_at | No | |
| graduation | No | |
| launch_minute | No | |
| curve_progress_pct | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly/idempotent/non-destructive, and the description goes well beyond them: pricing ($0.01/call, 10 free/day, x402 payment-required result), freshness (rebuilt every 6 hours), shrinkage behavior on thin samples, and error semantics (isError on invalid input or upstream failure, uncharged). This is unusually rich disclosure for a read-only tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loaded with the core purpose and method, followed by freshness, when-to-use, price, and errors. The sentences are dense but each earns its place; slightly long, though nothing is truly redundant.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Output schema exists, so return-value explanation is unnecessary. The description still covers the modeling approach, data freshness, cost, error behavior, and routing to alternatives, leaving no material gap for calling the tool correctly.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Only one parameter with 100% schema description coverage, so the schema fully documents the required mint and its Base58 format. The description's bucket details (unique buyers, buy SOL, dev buy over the first 60s) describe the model inputs rather than parameter usage, adding no semantics beyond the schema. Baseline 3 applies.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific resource (pump.fun launch graduation probability) and a specific method (empirical rates from comparables with lift over the ~3% base rate). It distinguishes itself from siblings by scoping to 'graduation probability only' and naming launch_scan as the broader alternative. An agent can tell this apart from pump_launches, pump_token, or pump_pulse without opening a schema.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicit 'When to use: For graduation probability only' with a named alternative (launch_scan on the MAD Synapse risk server) that bundles it with other checks. This is a genuine when/when-not routing statement, not implied guidance.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Live pump.fun heartbeat: trades/min, launches, graduations, graduation rate, hottest tokens in the last 15 min. Computed from our own live recording of every pump.fun bonding-curve trade. Use it to decide whether the launch market is hot or dead before scanning individual tokens. When to use: For the market right now; for trends over days use pump_history. Price: free. Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| as_of | No | |
| sol_usd | No | |
| coverage | No | |
| last_24h | No | |
| last_hour | No | |
| hottest_15m | No | |
| trades_per_min | No | |
| recent_graduations | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations cover the safety profile (readOnly, idempotent, non-destructive, open-world), and the description adds value beyond them: cost ('Price: free'), error behavior ('returns isError ... not charged'), and data provenance from live recording. This is behavioral context an agent cannot get from the structured fields.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loaded with the core purpose, then data scope, then usage routing, then price and error semantics. Every sentence carries information (metrics, provenance, routing, cost, failure mode) with no filler.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
An output schema exists, so return values need not be explained, yet the description still lists the headline metrics. Combined with cost and error disclosure for a zero-param read tool, nothing an agent needs to invoke it correctly is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The tool takes no parameters, so there is nothing to document and the baseline is 4. The description correctly does not invent parameter semantics; it instead enumerates the output metrics, which is appropriate for a zero-arg tool.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a concrete verb and resource: a live market heartbeat with named metrics (trades/min, launches, graduations, graduation rate, hot tokens over 15 min). It also discloses the data source ('our own live recording of every pump.fun bonding-curve trade'), which distinguishes it from siblings like pump_history and pump_launches.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Explicit routing: 'Use it to decide whether the launch market is hot or dead before scanning individual tokens' and 'When to use: For the market right now; for trends over days use pump_history.' Names both the situation and the alternative, leaving nothing to inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Full bonding-curve forensics for one pump.fun mint: flow, dev buys/sells, launch snipers, top curve holders, peak, graduation, red flags. Aggregates every recorded curve trade of the token: unique wallets, buy/sell SOL, first-60-second flow, whether the dev sold out, non-dev wallets that bought in the launch slots (bundles), top net holders on the curve, peak market cap and post-graduation liquidity. When to use: For raw curve forensics on one mint. For a single scored verdict use launch_scan (on the MAD Synapse · Wallets & Risk server, https://agent.maddegen.art/hub/mcp/risk); for a contract/liquidity safety grade (also after graduation) use token_risk (on the MAD Synapse · Wallets & Risk server, https://agent.maddegen.art/hub/mcp/risk); for the creator's track record use pump_dev. Price: $0.01 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| mint | Yes | pump.fun token mint. Base58 Solana address, 32-44 chars. |
| Name | Required | Description |
|---|---|---|
| dev | No | |
| flow | No | |
| mint | No | |
| name | No | |
| note | No | |
| curve | No | |
| flags | No | |
| found | No | |
| mayhem | No | |
| symbol | No | |
| creator | No | |
| socials | No | |
| swap_url | No | |
| first_60s | No | |
| created_at | No | |
| token_program | No | |
| launch_snipers | No | |
| post_graduation | No | |
| top_curve_holders | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already cover readOnly/idempotent/openWorld safety, and the description adds real behavioral context beyond them: $0.01 per call, 10 free/day, x402 payment-required result shape, and that isError responses for invalid input or upstream failures are not charged. It doesn't describe latency or result size, but the cost/error disclosures are substantive.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Dense but front-loaded: the forensic scope and aggregation list come first, routing guidance second, cost/errors last. The aggregation enumeration is long-ish but each item is a distinct output category an agent needs; little dead weight.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Despite an output schema existing, the description still tells the agent what body of data it returns and what it excludes; combined with pricing, error semantics, and sibling routing, an agent has everything needed to call it correctly or choose otherwise.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Single parameter with 100% schema description coverage (base58 mint format, 32-44 chars) documented in the schema itself. The description adds no format, validation, or lookup semantics beyond the schema, so baseline 3 is appropriate.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific verb+resource+scope: 'Full bonding-curve forensics for one pump.fun mint,' then enumerates exactly what is aggregated (curve trades, dev buys/sells, snipers, holders, peak, graduation). It explicitly names sibling tools it is not (launch_scan, token_risk, pump_dev), so an agent can route without opening schemas.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Has an explicit 'When to use' clause contrasting this tool against three named alternatives on specific conditions (single scored verdict vs. raw forensics; contract/liquidity safety; creator track record), including the server URL for the alternatives. No inference required.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Any wallet's pump.fun trading record: realized PnL, win rate, hold time, frequency (bot or human), tokens it launched. Analyses up to the last 20,000 curve trades of the wallet. Use it to judge whether a wallet is smart money, a sniper bot, a serial dumper or a dev. When to use: For a trader's pump.fun performance. For a quick deal/no-deal verdict on any wallet use counterparty_check (on the MAD Synapse · Wallets & Risk server, https://agent.maddegen.art/hub/mcp/risk); to test whether copying it pays use copy_backtest. Price: $0.01 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | How many of the wallet's most recent recorded trades to analyse. Range 100-20000. Default 20000. | |
| wallet | Yes | wallet address. Base58 Solana address, 32-44 chars. |
| Name | Required | Description |
|---|---|---|
| note | No | |
| found | No | |
| label | No | |
| wallet | No | |
| window | No | |
| activity | No | |
| recent_tokens | No | |
| classification | No | |
| created_tokens | No | |
| closed_positions | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations only cover the safety profile (readOnly, idempotent, non-destructive); the description goes well beyond by disclosing the analysis window (last 20,000 curve trades), cost and quota ($0.01/call, 10 free/day, x402 payment-required result), and error semantics (isError, not charged). It also tells the agent how to interpret results (smart money, sniper bot, dumper, dev).
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
Front-loaded with the core capability and output list, then progressively adds usage, pricing, and error behavior. Dense but nearly every clause carries operational value; the long parenthetical URL is the one mildly bulky element.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With an output schema present, return values need not be enumerated, and the description still covers scope, cost, failure behavior, and interpretation guidance. Nothing an agent needs to invoke it correctly or budget calls is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so the baseline is 3. The description adds meaning the schema lacks by clarifying that the limit counts 'curve trades' and that analysis is capped at the last 20,000 — a semantic nuance beyond the bare 'how many recent trades' schema text.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
States a specific resource and scope: 'Any wallet's pump.fun trading record' followed by the exact outputs (realized PnL, win rate, hold time, frequency, tokens launched). It also differentiates itself from siblings by naming counterparty_check and copy_backtest and describing when each is preferred.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
Gives an explicit 'When to use' clause plus two named alternatives with the conditions that select them (deal/no-deal verdict vs. copy-trade viability). It even flags that counterparty_check lives on a different server with its URL, so routing decisions are unambiguous.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
The pump.fun wallets with the best realized PnL in the last 24 h — and the tokens those wallets are buying right now. Leaderboard of wallets by realized profit on fully-closed curve positions (min 8 closed, rebuilt every 30 min), sortable by PnL, win rate or ROI, plus a live feed of what the top 50 bought in the last N minutes, ranked by how many of them piled in. When to use: To find wallets worth watching; then copy_backtest to test copying one. Price: $0.02 per call (10 free/day; after that a payment-required result lists x402 options). Errors: returns isError with a message for invalid input or an upstream failure (not charged).
| Name | Required | Description | Default |
|---|---|---|---|
| sort | No | Rank by realized PnL in SOL ("pnl"), win rate ("win_rate") or return on SOL spent ("roi"). Default "pnl". | pnl |
| limit | No | How many top wallets to return. Range 1-50. Default 20. | |
| window_min | No | look-back for buying_now. Range 1-120. Default 15. | |
| exclude_bots | No | drop sub-20-second and high-frequency wallets. Default false. |
| Name | Required | Description |
|---|---|---|
| note | No | |
| sort | No | |
| window | No | |
| wallets | No | |
| buying_now | No | |
| leaderboard_built_at | No |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
Annotations already declare readOnly/idempotent/non-destructive, but the description goes well beyond by disclosing the $0.02-per-call price, the 10 free/day quota, the x402 payment-required path, the 30-minute rebuild cadence, and that invalid input or upstream failures return isError uncharged. It does not, however, describe the freshness/latency of the live feed beyond the window parameter. Strong behavioral disclosure.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The core output is front-loaded in the first sentence before the sortable/feed details, and the when-to-use, price, and error clauses are ordered usefully. It is dense and multi-clause, but nearly every clause adds a distinct fact rather than padding.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
With an output schema present, return values need not be re-explained, and the description still covers pricing, quota, error behavior, refresh cadence, ranking basis, and the natural follow-up tool. Nothing an agent needs to call it correctly is missing.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema coverage is 100%, so the baseline would be 3, but the description adds meaning by tying window_min to the buying_now feed ('last N minutes'), limit to the leaderboard ('top 50'), and sort to PnL/win-rate/ROI. It clarifies which param drives which half of the response, which the schema alone does not.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description states a precise verb+resource (a wallet leaderboard ranked by realized PnL plus a live buying feed) and immediately scopes it: 'fully-closed curve positions (min 8 closed, rebuilt every 30 min)'. It is clearly distinguishable from siblings like pump_wallet and copy_backtest.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
It gives an explicit when-to-use ('To find wallets worth watching') and routes the agent to the next step with a named alternative ('then copy_backtest to test copying one'). Nothing about tool selection is left to inference.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
copy_backtestdlmm_pooldlmm_positionsdlmm_whalespump_devpump_historypump_launchespump_oddspump_pulsepump_tokenpump_walletsmart_moneySolana + pump.fun intel for agents: launch verdicts, token risk, dev/wallet records, smart money.
Solana DeFi MCP: Meteora DLMM brief+detail; Solana DEX scout. Commerce optional.
Rug pull risk and on-chain forensics for tokens on Solana, Ethereum, Base and Robinhood.
Full Solana DeFi coverage: launchpads, tokens, trades, and wallets, decoded at scale.
Add one secure layer between your agents and this server.
