Loading ToolBench

Reading the current corpus…

MCPServer MCP server · D · ToolBench

MCPServer

A Python MCP server providing browser search, document rendering, PDF extraction, and office document conversion tools with async database persistence and caching.

2026-07-28+STREAMABLE_HTTP2 tools

58/100

Definition Quality67

Protocol Readiness75

Supportability33

Spec posture

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

Static source inference · medium confidence · detected: Logging

Deprecated protocol patterns detected

Summary

Two tools with complete input schemas and descriptions, but descriptions are generic and lack LLM-optimized guidance. Both tools follow verb_noun naming (browser_render_*). Schemas include proper types and parameter descriptions. However, descriptions lack context on WHEN to use each tool, what distinguishes them, and actionable error recovery. Output schemas are not documented. No error handling guidance. Parameters lack constraints (e.g., DPI range, page number validation). Missing idempotency hints and composition guidance.

Tools (2)

browser_render_docxread onlysource verified67/100

Render specified pages (or all pages) of a local Word document (.docx) into high-fidelity PNG preview images. Returns network-compliant Resource URIs and inline visual previews.

browser_render_pptxread onlysource verified67/100

Render specified slides (or all slides) of a local PowerPoint presentation (.pptx) into high-fidelity PNG preview images. Returns network-compliant Resource URIs and inline visual previews.

Issues (5)

HIGHtool-description ↗

Descriptions lack LLM-optimized guidance. Both tools describe WHAT they do but not WHEN to use them or how they differ. Generic descriptions like 'Render specified slides' do not help LLMs choose between tools or understand prerequisites.

browser_render_docxbrowser_render_pptx

HIGHresponse-shaper ↗

Output schemas are not documented. LLMs cannot plan downstream operations or extract required fields (e.g., image URIs, page counts, error details). Responses are opaque.

browser_render_docxbrowser_render_pptx

MEDIUMconstrained-input ↗

Parameters lack constraints and validation guidance. 'dpi' has no min/max (unbounded integers invite absurd values). 'pages' and 'slides' arrays lack length limits or index range documentation. No guidance on what happens if indices are out of bounds.

browser_render_docxbrowser_render_pptx

MEDIUM

Recommendations

Expand descriptions to 100-200 characters. Include WHEN to use each tool (e.g., 'Use this to preview document layouts before sharing') and distinguish from similar tools. Add dependency hints (e.g., 'Requires a valid .docx file path').
Document output schema explicitly. Specify returned fields: image_uri (string), page_number (integer), width (integer), height (integer), error (string, optional). Include example response structure.
Add parameter constraints: dpi (integer, 72-600, default 150), pages/slides (array of integers, 1-indexed, max 1000 items). Document behavior when indices exceed document length.
Implement actionable error messages. Instead of 'Invalid page', return 'Page 10 not found. Document has 5 pages. Valid pages: 1-5.' This enables LLM self-correction.
Add tool annotations in schema: readOnlyHint=true, idempotentHint=true. This signals to LLMs that retries are safe and no state is modified.

Spec posture evidence

Inferred effective spec: 2026-07-28+.

Relies on Logging (deprecated) - log to stderr or use OpenTelemetry

Score history

Overall score trend

First recorded score · v2 rubric

58/100

Scored	Grade	Overall	Spec posture	Rubric
2026-09-23	D	58	2026-07-28+	v2