Loading ToolBench

Reading the current corpus…

mcp-server-convert MCP server · D · ToolBench

mcp-server-convert

MCP server for converting documents to Markdown. Supports PDF, DOCX, HTML, EPUB, CSV, JSON, XML, XLSX, PPTX, and plain text.

2026-07-28+STDIO6 tools

55/100

Definition Quality72

Protocol Readiness50

Supportability31

Spec posture

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

Static source inference · medium confidence

Current-spec patterns detected

Summary

Server has 6 tools with complete schemas and descriptions, but quality is uneven. Tool names follow verb_noun convention (convert_*, extract_*, list_*), which is good. Descriptions are present but generic (avg ~80 chars, below 194-char baseline). Parameters have types and descriptions, but lack constraints (enums, ranges, patterns). No output schemas documented. Error handling returns plain text without recovery guidance. No tool annotations (readOnlyHint, etc.). Batch and directory tools lack pagination despite potentially returning many results.

Tools (6)

batch_convertread onlysource verified68/100

Convert multiple files to Markdown at once.

convert_directoryread onlysource verified67/100

Convert all supported files in a directory to Markdown.

convert_fileread onlysource verified77/100

Convert a local file to Markdown. Supports PDF, DOCX, HTML, EPUB, CSV, JSON, XML, XLSX, PPTX, and plain text.

convert_urlread onlysource verified75/100

Fetch a URL and convert its HTML content to Markdown.

extract_metadataread onlysource verified75/100

Extract metadata from a file (size, type, dates, etc.) without full conversion.

list_supported_formatsread onlysource verified67/100

List all supported file formats and their conversion methods.

Issues (5)

HIGHtool ↗

No output schemas documented. LLMs cannot predict response structure (fields, types, nesting). Downstream tool chaining and data extraction become error-prone.

convert_fileconvert_urlbatch_convertconvert_directoryextract_metadatalist_supported_formats

HIGHpaginated-result ↗

batch_convert and convert_directory lack pagination. No limit enforcement visible in code; max_files=20 default may silently truncate results without indicating more exist. No next_cursor or total_count returned.

batch_convertconvert_directory

HIGHrecovery-guide ↗

Error responses are plain text without recovery guidance. E.g., 'Error converting file: {e}' gives LLM no actionable next step. Should indicate: is this retryable? Should user check file permissions? Is format unsupported?

convert_fileconvert_urlbatch_convertconvert_directoryextract_metadata

Recommendations

Add output schemas to all tools. Example for convert_file: {type: 'object', properties: {content: {type: 'string', description: 'Markdown-formatted file content'}, truncated: {type: 'boolean'}, original_size: {type: 'integer'}}}
Implement pagination for batch_convert and convert_directory. Return {results: [{file_path, content, success, error}], total_count, next_cursor} and accept cursor/offset + limit params.
Enhance error messages with recovery hints. E.g., 'Error: File not found at {file_path}. Verify path is absolute and file exists.' or 'Error: Unsupported format .xyz. Supported: PDF, DOCX, HTML, EPUB, CSV, JSON, XML, XLSX, PPTX, TXT'
Add readOnlyHint: true to all tool schemas to signal safety to agents and clients.
Document parameter constraints in descriptions. E.g., max_length: 'Maximum output length in characters (1 - 1000000, default: 50000)' and max_files: 'Maximum files to process (1 - 100, default: 20)'

Score history

Overall score trend

First recorded score · v2 rubric

55/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	D	55	2026-07-28+	v2