Loading ToolBench

Reading the current corpus…

mcp-hono-vercel-template MCP server · C · ToolBench

← Browse

mcp-hono-vercel-template

github.com/cris-achiardi/mcp-hono-vercel-template

Template for building a remote MCP server with Hono, deployed on Vercel

2026-07-28+STREAMABLE_HTTP4 tools

67/100

Definition Quality82

Protocol Readiness75

Supportability35

Spec posture

Inferred effective spec 2026-07-28+·current baseline 2026-07-28

Static source inference · medium confidence · detected: Dynamic Client Registration, stateful session

Deprecated protocol patterns detected

Summary

Four tools with complete schemas and descriptions. Naming follows verb_noun convention (echo, convert-length, search-items, get-item). Descriptions are clear and actionable (avg ~120 chars). All parameters have types and descriptions. Input schemas use JSON Schema with constraints (minLength, minimum, maximum, default). However, output schemas are not documented, LLMs cannot predict response structure. Error handling is minimal; no recovery guidance. Tool composition is sound (single responsibility, chainable IDs). Missing tool annotations (readOnlyHint, idempotentHint) despite all tools being read-only.

Tools (4)

convert-lengthread onlysource verified83/100

Convert a length between units (mm, cm, m, km, in, ft, mi). Returns the converted value plus the exact conversion factor used. Prefer this over calculating by hand.

echoread onlysource verified83/100

Echo a message back, optionally repeated. Use to verify the MCP connection is alive and that arguments arrive correctly.

get-itemread onlysource verified78/100

Fetch a single row from "items" by its id. Use after search-items to get full detail.

search-itemsread onlysource verified82/100

Search rows in the "items" table by name, newest first. Returns matching rows with all columns.

Issues (5)

HIGHtool ↗

Output schemas not documented. LLMs cannot predict response structure (fields, types, nested objects). Agents must infer from trial-and-error.

echoconvert-lengthsearch-itemsget-item

MEDIUMtool ↗

Tool annotations missing. All four tools are read-only and idempotent, but lack readOnlyHint and idempotentHint in schema. Agents cannot distinguish safe tools from destructive ones.

echoconvert-lengthsearch-itemsget-item

MEDIUMrecovery-guide ↗

Error handling lacks recovery guidance. No documented error cases, retryability classification, or actionable error messages. Agents cannot self-correct on failure.

search-itemsget-item

MEDIUMpaginated-result ↗

search-items returns 'newest first' but no pagination cursor or total count documented. Large result sets risk context window exhaustion.

search-items

Recommendations

Document output schemas for all tools. Example for echo: {type: 'object', properties: {result: {type: 'string', description: 'The echoed message'}}}. For search-items: {type: 'object', properties: {items: {type: 'array', items: {type: 'object', properties: {id, name, ...}}}, total: {type: 'number'}}}.
Add tool annotations to schema. Mark all four tools with readOnlyHint: true and idempotentHint: true in the tool definition. This signals to agents that retries are safe.
Add error handling documentation. For search-items and get-item, document: 'Returns empty array if no matches (not an error). Returns 400 if query exceeds 1000 chars. Returns 500 if database unavailable (retryable).'
Enhance search-items pagination. Add 'offset' parameter (default 0) and return {items: [...], total: N, hasMore: boolean} so agents can iterate without guessing.
Clarify convert-length precision use case. Update description: 'Decimal places in result (0-10). Use 2-4 for display, 8+ for calculations requiring high precision.'

Spec posture evidence

Inferred effective spec: 2026-07-28+.

Relies on Stateful initialize / Mcp-Session-Id (removed; protocol is stateless) - make each request self-contained

Score history

Overall score trend

First recorded score · v2 rubric

67/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	C	67	2026-07-28+	v2