BC pdf-to-latex-mineru
Convert PDF documents to LaTeX source code using MinerU AI extraction. Designed for researchers, academics, and scientists who need to re-edit, re-typeset, or recover LaTeX markup from published papers, theses, and technical reports. Use when asked to: convert a PDF paper to LaTeX, extract LaTeX from an academic PDF, edit a PDF in LaTeX, re-typeset an arXiv paper, recover LaTeX source from PDF, turn a document into editable LaTeX, get equations out of a PDF, extract math formulas from document, convert research paper to LaTeX. Handles complex academic layouts: mathematical equations, multi-column text, tables, figures, and scientific notation. Supports local PDF files and direct URLs including arXiv links. Use --model vlm for high-accuracy extraction of math-heavy or multi-column documents; use pipeline mode for guaranteed structural fidelity. Solves problems like: I have a PDF but need the LaTeX source, I need to modify a paper but only have the PDF, I want to reuse equations from a published paper, how do I make a scanned paper editable in LaTeX. Powered by MinerU from OpenDataLab (Shanghai AI Lab) - open-source, high-quality PDF understanding engine. Requires MINERU_TOKEN. 将PDF学术论文转换为LaTeX源码。支持公式提取、多栏排版识别、表格还原， 适用于论文重排版、学术编辑、arXiv论文转换等场景。
As a process C 59/100 · Has gaps — weak spots: when it triggers, inputs and preconditions, failures and branches
How to improve
- Shorten the description to 1024 characters.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 2. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 1262 chars, limit 1024 - note
frontmatter-keyunknown frontmatter key "homepage"
Process rating: all ten parameters 59/100
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Failures and branches. Linear process with no failure handling
- 0Progress reporting. Says nothing while it works
- 20When it triggers. No condition that starts the skill
- 30Running it twice. 1 mutating operations with no state check
- 60Result and completion. Output format stated, no completion criterion
- 100Tools and files. No external tools needed
- 100Steps. 9 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 318 tokens
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 1261: 120–800 characters recommended
- +2Single-language instructions
- +4Structure: 6 headings
- +3Step-by-step instructions: 9 items
- +3Output format is stated explicitly
- +4Has examples (3 code blocks)
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 59.