Name: markitdown
Rating: 67.2 (1 reviews)
Author: K-Dense-AI

markitdown

Convert heterogeneous documents and selected URIs to Markdown with Microsoft MarkItDown for text analysis, search, and LLM/RAG ingestion. Covers safe local conversion, streams, Office/PDF/data formats, batch workflows, plugins, vision OCR, Azure extraction, and the official MCP server.

Quality

80%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

The canonical home for this skill is markitdown in K-Dense-AI/scientific-agent-skills

Quality

Content

86%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is well-structured and highly actionable, with executable code, a routing table, and exemplary progressive disclosure through a reference-file index. The main opportunity is tightening the Overview prose and making the batch validation feedback loop explicit.

Suggestions

Trim the Overview's explanatory sentences about what MarkItDown is and its output philosophy to keep the body fully lean, or move that framing into a reference file.

Add an explicit validate→fix→retry feedback loop for batch conversion (e.g., inspect the manifest for failures, re-run failed inputs with corrected flags) to push workflow clarity to the top anchor.

Consider a one-line 'Use when...' trigger note near the top of the body to mirror the description's missing trigger guidance and aid skill selection.

Dimension	Reasoning	Score
Conciseness	Largely lean and table-driven with terse rules and executable snippets that assume competence, but the Overview contains a few explanatory sentences ('MarkItDown is Microsoft's lightweight Python utility...', 'Its output is designed primarily for...') that could be trimmed without losing clarity.	4 / 5
Actionability	Fully executable, copy-paste-ready code (convert_local, convert_stream with StreamInfo), concrete CLI invocations with flags and pinned versions, and complete batch/literature script commands — covering the common cases specifically and completely.	5 / 5
Workflow Clarity	A clear sequenced Quality Checks checklist with explicit 'do not infer success' guidance and documented batch flag behavior provides most checkpoints, but the batch path lacks an inline validate→fix→retry feedback loop, keeping it just below the top anchor.	4 / 5
Progressive Disclosure	Clear overview with a 'Reference Files' table mapping each of 7 one-level-deep reference files (all verified present) to explicit 'Read when' conditions, with bulk detail pushed to references and easy navigation.	5 / 5
	Total	18 / 20 Passed

Description

75%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific, distinctive, and clearly states what the skill does across a broad feature surface, but it omits an explicit 'when to use' trigger clause, which caps its completeness. Adding a concise 'Use when...' sentence would lift completeness and trigger-term quality.

Suggestions

Add an explicit 'Use when...' clause (e.g., 'Use when converting PDFs, Office docs, HTML, or URIs to Markdown for LLM/RAG ingestion, search, or text analysis.') to satisfy the completeness 'when' requirement.

Surface a few more natural trigger synonyms and file extensions users actually say (e.g., 'extract text', '.docx', '.pptx', '.xlsx') to push trigger-term quality to comprehensive.

Keep the existing concrete capability list as-is — it already anchors specificity and distinctiveness well.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities across the tool's scope — 'Convert heterogeneous documents and selected URIs to Markdown', 'safe local conversion, streams, Office/PDF/data formats, batch workflows, plugins, vision OCR, Azure extraction, and the official MCP server' — giving comprehensive coverage rather than generic language.	5 / 5
Completeness	The 'what' is clear and thorough, but there is no 'Use when...' clause or equivalent explicit trigger guidance, so per the rubric guideline a missing trigger clause caps completeness at 3.	3 / 5
Trigger Term Quality	Strong natural keyword coverage ('documents', 'Markdown', 'Office/PDF', 'text analysis, search, LLM/RAG ingestion', 'batch workflows', 'OCR', 'Azure'), but a few common synonyms a user might say (e.g., 'extract text', explicit '.docx') are not stated, so it sits just below the comprehensive anchor.	4 / 5
Distinctiveness Conflict Risk	Tied to a specific named tool (MarkItDown) with a sharply defined niche (document/URI → Markdown for LLM/RAG ingestion), giving a clear niche with minimal conflict risk.	5 / 5
	Total	17 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: K-Dense-AI/claude-scientific-writer
Path: .claude/skills/markitdown/SKILL.md
Commit: 14ee3e3

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.