Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. Tool names follow verb_noun convention (get_, extract_), which is good. However, descriptions lack actionable context for LLM selection. The extract_clean_markdown tool has a critical missing parameter description (paymentTxHash), and neither tool documents output structure. Error handling is present in code but not reflected in schema. STDIO transport caps overall potential.

• Add explicit output schema to both tools. Example for get_payment_info: 'Returns object with fields: protocol (string), network (string), chainId (number), token (string), tokenAddress (string), recipient (string), priceUsdc (number), instructions (string).'
• Expand extract_clean_markdown description to 150-200 chars: 'Extract clean, LLM-optimized Markdown from any web URL on Base L2. Requires prior 0.05 USDC payment transaction. Returns markdown content, word count, and extraction metadata. Call get_payment_info first to get recipient address.'
• Add description to paymentTxHash parameter: 'The EVM transaction hash (0x + 64 hex chars) of the 0.05 USDC transfer to the agent on Base L2. Must be a successful, non-replayed transaction.'
• Document error cases in tool descriptions or add an errors field to schema. Example: 'Errors: 402 Payment Required (no tx hash provided), 400 Invalid tx hash format, 402 Insufficient payment, 409 Replay attack detected, 400 Transaction failed on Base L2.'
• Consider adding a 'prerequisites' or 'dependencies' field to extract_clean_markdown description: 'Prerequisite: Call get_payment_info to retrieve recipient address, then send 0.05 USDC on Base L2 mainnet (chainId 8453) to that address before calling this tool.'