Loading ToolBench

Reading the current corpus…

citadel-cli MCP server · F · ToolBench

← Browse

citadel-cli

github.com/Citadel-Inc/citadel-cli

Command-line interface for Citadel platform with MCP (Model Context Protocol) support for tools, resources, and prompts

<=2025-11-25STDIO4 tools

29/100

Definition Quality19

Protocol Readiness47

Supportability33

Spec posture

Inferred effective spec <=2025-11-25·current baseline 2026-07-28

Static source inference · medium confidence · detected: stateful session

Deprecated protocol patterns detected

Summary

citadel-cli exposes 4 MCP tools with minimal definition quality. Tool names lack action verbs (mcp/tools, mcp/call, mcp/resources, mcp/prompts are namespace-prefixed but not verb-first). Descriptions are present but generic (10-50 chars, below the 50-200 char LLM-optimized baseline). Only mcp/call has a visible input schema with two parameters (tool_name, arguments), but both lack detailed descriptions. mcp/tools, mcp/resources, and mcp/prompts have no visible input schemas in the provided code. No output schemas documented. Error handling is marked present but not visible in the code sample. No tool annotations (readOnlyHint/destructiveHint) despite clear risk classifications (READ_ONLY vs WRITE). Parameter descriptions are minimal or absent.

Tools (4)

mcp/callwriteauth35/100

Call an MCP tool with specified arguments

mcp/promptsread onlyauth20/100

List available MCP prompts from a connected MCP server

mcp/resourcesread onlyauth20/100

List available MCP resources from a connected MCP server

mcp/toolsread onlyauth20/100

List available MCP tools from a connected MCP server

Issues (5)

CRITICALtool ↗

No input schemas visible for 3 of 4 tools (mcp/tools, mcp/resources, mcp/prompts).

mcp/toolsmcp/resourcesmcp/prompts

HIGHtool ↗

Tool names do not start with action verbs. 'mcp/call' is acceptable but 'mcp/tools', 'mcp/resources', 'mcp/prompts' are nouns, not verbs. Should be 'list_mcp_tools', 'list_mcp_resources', 'list_mcp_prompts'.

mcp/toolsmcp/resourcesmcp/prompts

HIGHtool-description ↗

Descriptions are 10-50 characters, well below the 50-200 char LLM-optimized baseline. E.g. 'List available MCP tools from a connected MCP server' (52 chars) lacks context on WHEN to call it or what it returns.

mcp/toolsmcp/callmcp/resourcesmcp/prompts

HIGHtool-description ↗

mcp/call parameters 'tool_name' and 'arguments' lack detailed descriptions. 'arguments' is typed as object but no guidance on structure, required fields, or format.

Recommendations

Rename tools to verb-first: 'list_mcp_tools', 'call_mcp_tool', 'list_mcp_resources', 'list_mcp_prompts'. This helps LLMs infer intent from the name alone.
Expand descriptions to 100-150 chars. E.g. 'List all available tools from the connected MCP server. Call this first to discover what operations are available before invoking call_mcp_tool.'
Add explicit input schemas for mcp/tools, mcp/resources, mcp/prompts (even if empty, document as {}). Document output schemas (e.g., tools returns {tools: [{name, description, inputSchema}]})
Add tool annotations: readOnlyHint=true for list_* tools; destructiveHint=true for call_mcp_tool if it can modify state.
Enhance mcp/call parameter descriptions: 'tool_name (string): Name of the MCP tool to invoke, e.g. "get_user". Call list_mcp_tools first to discover available names.' and 'arguments (object): Tool-specific parameters as a JSON object. Structure depends on the tool's inputSchema.'

Spec posture evidence

Inferred effective spec: <=2025-11-25.

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

29/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	29	<=2025-11-25	v2