ankr-mcp provides 21 well-named, read-only blockchain query tools with complete JSON schemas generated from Zod validators. All tools have descriptions (avg ~80 chars) and typed parameters. However, descriptions are terse and lack LLM-optimized guidance on WHEN to use each tool, dependencies between related tools, or recovery hints for common failures. Parameter descriptions are minimal (often 1-2 sentences). Output schemas are not documented in the tool definitions, LLMs cannot predict response structure. Error handling is absent from tool definitions; no guidance on retryability or user-fixable vs fatal errors. No tool annotations (readOnlyHint, idempotentHint) despite all being read-only. Composition is strong: tools are single-purpose and chainable (e.g., get_token_balances → get_token_price). Naming follows verb_noun convention consistently.

• Add output schema documentation to each tool. Document the response structure (e.g., 'Returns {balances: [{token: string, amount: number, usdValue: number}], total: number}') so LLMs can plan downstream calls and validate chaining.
• Expand tool descriptions to 100-150 chars with LLM-optimized guidance: WHAT it does, WHEN to use it vs similar tools, and any prerequisites. Example: 'Gets token balances for a wallet. Use this BEFORE get_token_price to identify which tokens to price. Supports ENS names.'
• Add error handling guidance to descriptions. Document common failures: 'If address not found, try search_addresses(). If Premium API error, upgrade account. Retryable on timeout.'
• Add readOnlyHint: true to all tool definitions in the ListToolsResponse. This signals to clients that these tools are safe to cache and retry without side effects.
• Enhance parameter descriptions with format constraints and valid ranges. Example: 'blockchain (string, one of: eth, bsc, polygon, base, arbitrum), the blockchain network to query' instead of just 'The blockchain network'.