Skip to content
Get packs

LiteParse Local Document Parser

Extract text, layout boxes, OCR and page images from PDFs and Office files locally

Source Download skill
Repository
K-Dense-AI/scientific-agent-skills
Stars
55,780
Licence
MIT
Files
7 files

What it does

Lets the agent read PDFs, Word and Excel files, and images on your own machine and pull out their text along with where each piece sits on the page. It can run OCR on scanned pages, output structured JSON or Markdown for search and retrieval systems, and render pages as PNG images so a multimodal agent can see charts and figures. It can also batch-process a whole folder of documents.

Great for

  • Researchers who need to ingest folders of papers or supplementary PDFs into a local RAG pipeline
  • Developers building retrieval or citation features that need text with bounding boxes
  • Anyone who needs to convert scanned documents to text without uploading them to a cloud service

Skip it if

  • Dense tables, handwriting or production cloud pipelines, where the skill points to LlamaParse, a separate cloud service that requires its own sign-up
  • Merging, splitting or editing PDFs, which needs a PDF manipulation library instead
  • Sensitive documents if you enable the optional HTTP OCR server, because it sends document images to that server

Try asking

  • “Parse paper.pdf to JSON with bounding boxes for each text item”
  • “Batch-parse everything in ./papers into ./parsed as JSON, recursively”
  • “Render pages 1, 3 and 5 of this report as PNG screenshots”
  • “OCR this scanned contract and return layout-preserved text”
Step-by-step guide: how to use LiteParse Local Document ParserInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 7 original files, the MIT licence and an attribution note.

    liteparse.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip liteparse.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip liteparse.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Use --no-ocr for text-native PDFs to speed up parsing
  • Restrict batch input with --extension or use the wrapper script to avoid output filename collisions for same-named files
  • Pre-install Tesseract language data if you want to avoid network downloads during parsing

What's inside

  • references/api_reference.md
  • references/choosing_a_parser.md
  • references/cli_reference.md
  • references/ocr_and_formats.md
  • references/output_formats.md
  • scripts/batch_parse_dir.py
  • SKILL.md

Redistributed unmodified from K-Dense-AI/scientific-agent-skills under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.