Model Context Protocol (MCP) server for TON Access - provides unthrottled anonymous RPC access to the TON blockchain via a robust decentralized network of nodes
Server provides 6 tools with basic schemas and descriptions. All tools have descriptions (10-50 chars, below ideal 50-200 char range). Input schemas are present but minimal: most parameters lack individual descriptions. Tools 5-6 (ADNL endpoints) have empty input schemas ({}), violating schema completeness. No output schemas documented. Error handling absent, no recovery guidance or categorization. Tool naming follows verb_noun pattern (getHttpEndpoint, getAdnlProxyEndpoint) but descriptions are terse and lack context on when to use each variant. No parameter-level descriptions visible in schema definitions.
Get ADNL proxy endpoint information
Get multiple ADNL proxy endpoints
Get a single HTTP endpoint for TonCenter API v2
Get multiple HTTP endpoints for TonCenter API v2
Get a single HTTP endpoint for TonHub API v4
Get multiple HTTP endpoints for TonHub API v4
Tools 5-6 (getAdnlProxyEndpoint, getAdnlProxyEndpoints) have empty input schemas ({}).
All tool descriptions are under 50 characters and lack context on when to use each variant. E.g., 'Get a single HTTP endpoint for TonCenter API v2' does not explain the difference from getHttpEndpoints or when an LLM should choose one over the other.
No output schemas documented. LLMs cannot plan downstream calls or extract required fields (e.g., endpoint URL, node ID, health status) without knowing the response structure.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 37 | 2026-07-28+ | v2 |
Parameter descriptions missing in schema definitions. The 'config' object in tools 1-4 has a description ('Optional configuration') but nested properties (network, host, accessVersion, protocol) lack individual descriptions explaining their purpose and constraints.
No error handling or recovery guidance. Tools do not document what happens on invalid network, unsupported protocol, or unreachable endpoint. LLMs have no guidance on retry strategy or fallback actions.