Server has 10 tools with consistent naming (verb_noun pattern: get_*, all READ_ONLY). All tools have descriptions (avg ~80 chars, baseline 194 chars, below target). Input schemas present with Zod validation and descriptions for all parameters. However, output schemas are NOT documented, responses are returned as JSON text blobs without formal schema declarations. Error handling is basic (try/catch with error messages) but lacks recovery guidance or error classification. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite all being read-only. Parameter descriptions are adequate but lack format constraints (e.g., address format, network enum values). Missing pagination for tools that could return large results (e.g., get_supported_networks).

• Add formal output schema declarations to all tools. Document response structure (e.g., {chainId: number, blockNumber: string, rpcUrl: string}) so LLMs can plan downstream calls.
• Add tool annotations: set readOnlyHint=true on all 10 tools to signal safety and enable agent optimizations (e.g., caching, parallel execution).
• Convert 'network' parameter to enum constraint listing all supported networks (sei, sei-testnet, sei-devnet, etc.) instead of free-form string. Update parameter description to reference the enum.
• Add address format validation in parameter descriptions: 'EVM address (0x + 40 hex characters)' for address params. Consider adding regex pattern to schema.
• Enhance error messages with recovery guidance: 'Invalid address format. Expected 0x + 40 hex chars. Example: 0x1234567890abcdef...' and classify errors (retryable, user-fixable, fatal).