Loading ToolBench
MCP server for interacting with Pump.fun tokens on Solana blockchain. Enables token creation, buying, selling, and account management.
Server has 6 tools with explicit schemas and descriptions. All tools are properly registered with Zod schemas and descriptions present. However, descriptions are minimal (10-50 chars), parameter descriptions lack detail on constraints and formats, and output schemas are not documented. Error handling returns generic messages without recovery guidance. No tool annotations (readOnlyHint/destructiveHint) despite clear risk distinctions. Naming is verb-first and clear, but parameter descriptions fail to explain expected formats, ranges, or dependencies.
Buy a Pump.fun token
Create a new Pump.fun token
Get the SOL and token balances for an account
Get information about a Pump.fun token
List all accounts in the keys folder
Sell a Pump.fun token
Descriptions are too brief (10-50 chars). Rubric baseline is 194 chars. LLMs cannot determine when to select tools or understand prerequisites without fuller context.
Parameter descriptions lack constraint details. E.g., 'slippageBasisPoints' says '1% = 100' but does not specify valid range (0-10000?). 'initialBuyAmount' lacks minimum context. Rubric requires format, range, and allowed values in descriptions.
No output schemas documented. LLMs cannot plan downstream calls or extract required fields. Rubric requires documented return types for all tools.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 49 | 2026-07-28+ | v2 |
Error responses are generic ('Error getting token info: <message>'). No recovery guidance. Rubric requires actionable errors that tell LLM what to do next (e.g., 'Try search_tokens() with partial name').
No tool annotations (readOnlyHint/destructiveHint/idempotentHint). create-token, buy-token, sell-token are destructive/write operations but lack annotations. Rubric requires risk classification for agent safety.