Loading ToolBench

Reading the current corpus…

mcp-neurolora MCP server · F · ToolBench

← Browse

mcp-neurolora

github.com/shaneholloman/mcp-neurolora

An MCP server for collecting and documenting code from directories

2026-07-28+STDIO4 tools

46/100

Definition Quality53

Protocol Readiness50

Supportability33

Spec posture

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

Static source inference · medium confidence

Current-spec patterns detected

Summary

Four tools with mixed quality. All have descriptions and input schemas, but descriptions are generic and lack LLM-optimized guidance. Parameters mostly lack descriptions (critical gap). No output schemas documented. Error handling is minimal, no recovery guidance or actionable error messages. Tool names are verb-based but descriptions don't explain WHEN to use each tool or what distinguishes them. No tool annotations (readOnlyHint, destructiveHint). Schemas present but incomplete, many parameters undescribed.

Tools (4)

analyze_coderead onlyauthsource verified55/100

Analyze code using OpenAI API (requires your API key). The analysis may take a few minutes. So, wait please.

collect_codewritesource verified58/100

Collect all code from a directory into a single markdown file

create_github_issueswriteauthsource verified50/100

Create GitHub issues from analysis results. Requires GitHub token.

install_base_serverswritesource verified48/100

Install base MCP servers to the configuration

Issues (5)

CRITICALtool-description ↗

Parameter descriptions missing or minimal. 'owner', 'repo', 'issueNumbers' in create_github_issues lack descriptions. 'configPath' in install_base_servers has no description. LLMs cannot infer parameter meaning from names alone.

install_base_serverscreate_github_issues

HIGHresponse-shaper ↗

No output schemas documented. Tools return ToolResult with generic 'text' content, but LLMs need to know what fields/structure to expect for downstream chaining. E.g., does collect_code return file path, line count, token estimate?

collect_codeinstall_base_serversanalyze_codecreate_github_issues

HIGHtool-annotations ↗

No tool annotations (readOnlyHint, destructiveHint, idempotentHint). collect_code and install_base_servers are WRITE operations but lack destructiveHint. analyze_code is READ-only but lacks readOnlyHint. Agents cannot distinguish safe from risky tools.

collect_codeinstall_base_serversanalyze_codecreate_github_issues

Recommendations

Add descriptions to ALL parameters. E.g., 'configPath: Absolute path to MCP settings.json file (e.g., ~/.mcp/settings.json). Must be readable and writable.'
Document output schemas in tool descriptions. E.g., 'Returns {filePath: string, lineCount: number, tokenEstimate: number, duration_ms: number}'.
Add tool annotations: mark collect_code, install_base_servers, create_github_issues with destructiveHint=true; mark analyze_code with readOnlyHint=true.
Enhance descriptions with WHEN/WHY. E.g., 'Use collect_code to gather all source files into one markdown before calling analyze_code. Supports ignorePatterns like .gitignore.'
Add error recovery guidance. E.g., 'analyze_code requires OPENAI_API_KEY env var. If missing, ask user to set it. If rate-limited, retry after 60 seconds.'

Score history

Overall score trend

First recorded score · v2 rubric

46/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	F	46	2026-07-28+	v2