MadeOnSol · Agent Intelligence Gateway
One research layer for discovering opportunities, evaluating token risk, inspecting wallets and deployers, checking signal history, and following changes on Solana and Robinhood Chain. Structured evidence and explicit coverage limits — not another opaque trading recommendation.
Read-only intelligence · Existing subscriptions remain primary · Chain support follows verified source coverage
Each action uses supported MadeOnSol data sources instead of inventing a new index. Reports indicate source, freshness and where a factor could not be checked.
discover_opportunities
Rank observed KOL consensus and trending tokens, with launchpad context where available. This is research discovery, not a return prediction.
Coverage: Both chains for tracked KOL cohorts; near-bonded Solana sources are not RHC parity.
evaluate_token
Bring token risk, liquidity, early buyers, deployer history and tracked KOL evidence into one traceable report.
Coverage: Solana and Robinhood Chain; a missing source is explicitly partial, never assumed safe.
inspect_wallet
Investigate observed trades, cost-basis PnL, behavior and available reputation without mistaking FIFO-open positions for verified current holdings.
Coverage: Both chains, subject to tracked activity and available balance verification.
inspect_deployer
Connect a creator to indexed prior launches, outcomes, alerts and reputation. RHC market-cap milestones are not Solana bonding.
Coverage: Both chains; coverage and launchpad definitions differ.
evaluate_signal
Review historical signal outcomes, sample sizes, test windows and limitations rather than treating one successful alert as an edge.
Coverage: Solana Signal Scorecard; no RHC Scorecard parity is claimed.
changes_since
Consume source-scoped, cursor-paginated changes with provenance, replay-window limits and explicit gap errors.
Coverage: Seven curated KOL/deployer event families across Solana and RHC, not every trade or transfer.
Your subscription defines authorized, recurring API access. x402 funds individual eligible data requests. MCP exposes those capabilities to an agent client. They solve different problems and can be combined.
Production bots, terminals and recurring agents
Use a paid, subscriber-owned API key with existing tier limits and one compositional request per investigation. PRO, ULTRA, BUSINESS and Enterprise rights are enforced per underlying data source.
Compare subscriptionsOccasional research or wallet-funded agent tasks
Pay for eligible, individually priced public data endpoints without a recurring subscription. Not every composite Agent Gateway tool is automatically an x402 product; verify the live priced catalog and its 402 accepts[] before paying.
Explore current x402 dataClaude, Cursor and autonomous tool clients
MCP exposes tools to the agent. It is a connection method, not a third pricing plan: use an authorized API key or a supported keyless wallet-funded route. Existing package versions and local stdio mode remain valid until a separate public remote MCP transport is released.
Install current MCP serversRobinhood Chain data may be payable through USDG on RHC or USDC on Solana on eligible x402 routes. The response's data chain does not identify the asset used to settle the payment. Check the live offered legs and price.
An agent needs to distinguish unavailable inputs from a safe result. Gateway responses contain an evidence trail, source limitations and an explicit available, partial or unavailable state. Thechanges_sincecursor is source-scoped: if replay cannot be proven complete, the reader returns an error instead of silently skipping events.
No swap execution, no custody, no inferred all-chain coverage.
Illustrative response shape
{
"tool": "evaluate_token",
"chain": "solana",
"status": "partial",
"generated_at": "2026-10-09T10:00:00Z",
"evidence": [
{
"source": "token_risk",
"observed_at": null,
"status": "unavailable",
"limitations": [
"Source unavailable"
]
}
],
"warnings": [
"Do not infer safety from absent evidence"
],
"data": null
}Illustrative contract only, not an observed token assessment.
Two practical guides cover different decisions: how to structure a memecoin research workflow, and which access model makes sense when that workflow moves into production.
Developer tutorial
Walk through six source-backed intelligence checks, chain-specific outcomes, risk evidence, historical signals and cursor replay.
Read the agent tutorialIntegration & pricing
Compare tool transport, wallet-funded data payments and recurring subscriber rights, including RHC settlement rails and recovery.
Compare access modelsNo. The Gateway adds six intent-oriented research actions; the existing MCP packages continue to provide detailed endpoint tools. MCP is how an agent calls tools. It does not determine whether access is billed through a subscription or an eligible x402 transaction.
Not automatically. The standalone x402 catalog lists individually priced data calls. A Gateway action is only available via x402 when that exact composite route, price, network and paid-result recovery path have been explicitly released. Otherwise use an authorized subscriber API key.
No. Sources and token-launch semantics differ. For example, the current Signal Scorecard is Solana-oriented; RHC almost-bonded coverage is not assumed. Reports disclose unavailable and partial sources rather than inventing parity.
No. Research actions return data and evidence. Trading execution and user funds remain in your own application or agent. An optional x402 client may sign a data-payment authorization; that is not trading execution.
The subscription rights for redisplaying MadeOnSol data are different from personal research access. Commercial redistribution or a customer-facing terminal requires the appropriate BUSINESS or Enterprise agreement; check the current terms and plan details.
Connect an agent using the technical reference, choose a permitted subscriber tier or eligible wallet-funded data calls, and inspect every source limitation before making a decision.