Loading ToolBench

Reading the current corpus…

MCP-Client MCP server · F · ToolBench

MCP-Client

A Python client library for the Model Context Protocol (MCP) that enables communication with MCP servers via multiple transport types (stdio, HTTP, WebSocket, SSE)

2025-06-18+5 tools

12/100

Definition Quality0

Protocol Readiness0

Supportability39

Spec posture

Inferred effective spec 2025-06-18+·current baseline 2026-07-28

Static source inference · medium confidence · detected: Sampling, Roots, Logging, Dynamic Client Registration

Deprecated protocol patterns detected

Summary

This is a test/demo server, not a production MCP server. The 5 tools are defined only as bare JSON objects in tests/advanced_test_server.py with minimal structure. No input schemas are visible in the source code, tools list only name and description. The server is demonstrating MCP protocol features (sampling, elicitation, roots, notifications) but does NOT expose realistic, production-quality tool definitions. Tools appear to be triggers for testing client-side MCP features rather than functional tools for agent use.

Tools (5)

trigger_elicitationread onlysource verified

Ask client to elicit info

trigger_list_rootsread onlysource verified

Ask client for its roots

trigger_notificationsread onlysource verified

Send list_changed notifications

trigger_resource_updateread onlysource verified

Send resource update notification

trigger_samplingread onlysource verified

Ask client to sample a message

Issues (5)

CRITICALtool ↗

NO INPUT SCHEMAS VISIBLE. All 5 tools lack inputSchema definitions. Code shows only {"name": "...", "description": "..."} with no inputSchema key.

trigger_samplingtrigger_elicitationtrigger_list_rootstrigger_notificationstrigger_resource_update

CRITICALtool-description ↗

DESCRIPTIONS ARE TRIVIAL (10 - 25 chars). All 5 tools have descriptions under 50 characters with no detail on what the tool actually does, what parameters it accepts, when to call it, or what it returns. These descriptions provide almost no LLM guidance.

trigger_samplingtrigger_elicitationtrigger_list_rootstrigger_notificationstrigger_resource_update

HIGHtool ↗

NAMING LACKS CLARITY & ACTION VERBS. Tool names start with 'trigger_' (a generic action) rather than domain-specific verbs like 'sample_', 'elicit_', or 'list_'. 'trigger_sampling' is vague about what happens when called, does it request sampling from the client? Start the sampling process? Unclear naming confuses both LLMs and humans.

Recommendations

Add complete inputSchema (JSON Schema) to all tools. At minimum: {"type": "object", "properties": {...}, "required": [...]}. See src/mcp/types/tools.py Tool class, inputSchema is a required dict field.
Expand all descriptions to 50 - 200 characters explaining WHAT the tool does, WHEN to call it, and WHAT it returns. E.g. 'Request the MCP client to sample an LLM response via sampling/createMessage and return the client's generated message.' vs 'Ask client to sample a message'.
Rename tools from 'trigger_X' to action-verb names: 'create_sample_message', 'request_user_input', 'fetch_filesystem_roots', 'broadcast_notification_update', 'notify_resource_change'. These names convey intent.
Document parameter types and constraints. For each parameter, declare its type (string, integer, etc.), allowed values (enum), min/max bounds, and a description. E.g. 'maxTokens: integer, range 1 - 4096, default 100'.
Add output schemas to tool definitions. Document what fields the result.content array contains, the structure of each item, and any pagination or metadata fields.
For tools that take no parameters (like trigger_list_roots), explicitly declare inputSchema: {"type": "object", "properties": {}, "required": []}. Do not omit inputSchema.
Add error handling guidance. If a tool can fail (e.g., client timeout, invalid schema), document recovery steps. E.g. 'If the client does not respond within 30s, retry with a simpler message' vs silent failure.

Score history

Overall score trend

↓ 8 points across a rubric change (v1 → v2)

12/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-22	F	12	2025-06-18+	v2
2026-03-09	F	20	2024-11-05+	v1