MCP server for interacting with Monad testnet blockchain, including wallet operations, token dispersal, smart contract compilation and deployment
Server has 7 tools with basic schemas and descriptions, but significant gaps limit production readiness. All tools have descriptions (10-50 chars, mostly adequate) and input schemas with type constraints (regex patterns for addresses, array bounds). However, output schemas are completely undocumented, LLMs cannot predict return structure. Error handling is minimal (generic try-catch blocks returning unstructured text). No tool annotations (readOnlyHint/destructiveHint). Parameter descriptions lack actionable guidance on constraints and dependencies. The disperse-* tools accept arrays up to 100 items but lack pagination or result limits guidance. No recovery hints in error messages. Naming is verb-noun compliant but some tools conflate concerns (e.g., disperse-tokens handles both equal and custom amounts in one tool rather than separate variants).
Compile a Solidity smart contract
Deploy a compiled smart contract to Monad testnet
Disperse MON (native currency) to multiple addresses on Monad testnet
Disperse tokens to multiple addresses
Generate a TypeScript SDK for interacting with a deployed smart contract
Get MON balance for an address on Monad testnet
Transfer MON to a single address on Monad testnet
No output schemas documented. LLMs cannot predict return structure or chain tools. All 7 tools return unstructured text responses wrapped in {content: [{type: 'text', text: '...'}]} but this structure is not declared in tool definitions.
Error handling returns generic unstructured text. No recovery guidance, no error classification (retryable vs fatal), no invalid-value echoing. E.g., 'Failed to retrieve MON balance' gives LLM no actionable next step.
No tool annotations (readOnlyHint/destructiveHint/idempotentHint). Destructive tools (transfer-mon, disperse-mon, disperse-tokens, deploy-contract) lack explicit marking. LLMs cannot distinguish safe from risky operations.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 50 | 2026-07-28+ | v2 |
Parameter descriptions lack constraint details. E.g., 'amount' params describe units ('in ether') but omit min/max bounds, precision rules, or validation format. 'solcVersion' is optional but no guidance on default or valid format (e.g., '0.8.0' vs 'v0.8.0').
Disperse tools accept up to 100 recipients but no pagination, result limits, or batch guidance. No documentation of gas cost implications or transaction size limits. LLMs may attempt oversized batches.