Skip to main content
skill CX Codex

firecrawl-parse — Claude Code skill: local files to markdown

Turns a local PDF, DOCX or XLSX into markdown, writes a summary, or answers a question about the document. Built for file paths rather than URLs.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add firecrawl/cli --skill firecrawl-parse -a claude-code
Open on GitHub

What firecrawl-parse does

The skill wraps firecrawl parse and writes results under .firecrawl/. Input formats include PDF, DOCX, DOC, ODT, RTF, XLSX, XLS and HTML. Three modes: plain markdown conversion, a summary via -S, and a question answered with -Q.

Always pass -o. A parsed document can run into hundreds of KB, and streaming that to stdout floods the context window. Read the saved file in bounded chunks with head, grep or rg. Add .firecrawl/ to .gitignore.

Uploads cap at 50 MB per file. Expect roughly 1 credit per PDF page; HTML costs 1 flat. firecrawl credit-usage shows your balance but needs authentication. Quote any path that contains spaces.

What it helps with

  • convert a PDF to markdown for later search
  • pull a short summary out of a long report
  • ask a question about a document's contents
  • extract an XLSX table as text
  • check remaining credits before batch processing

Example prompts for the agent

  • Parse ./paper.pdf to markdown and save it as .firecrawl/paper.md
  • Summarize ./report.docx with the -S flag
  • Answer what the conclusions are in ./paper.pdf using -Q

How to install firecrawl-parse

1Skills CLI (Vercel)

npx skills add firecrawl/cli --skill firecrawl-parse -a claude-code

2Manually: put the files in place

~/.claude/skills/firecrawl-parse/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/firecrawl-parse/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about firecrawl-parse

How do I install firecrawl-parse in Claude Code?

Run npx skills add firecrawl/cli --skill firecrawl-parse -a claude-code. The skill lands at ~/.claude/skills/firecrawl-parse/SKILL.md.

What does firecrawl-parse do?

It converts a local file to markdown, summarizes it, or answers questions about it. Supported inputs include PDF, DOCX, XLSX and HTML.

How is firecrawl-parse different from firecrawl-scrape?

This one takes a local file path; firecrawl-scrape handles URLs. Reach for that skill when your input is a link.

Similar skills

All skills for Agent setup: memory, context, orchestration →

From the same set10

firecrawl-search This skill finds web sources, returns relevant excerpts, and can fetch full page content. Use it for research and structured data discovery. CX Codex +1 84,400installs firecrawl-scrape The skill reads a known page into markdown or JSON and runs discovered data-provider tools. Useful once the URL is already picked. CX Codex +1 83,500installs firecrawl-agent This skill navigates websites on its own and extracts structured data. Use it for multi-page catalogs, pricing data, and listings. CX Codex +1 81,300installs firecrawl-crawl Skill to automatically extract multiple pages from a website or section by following links. Use it to gather documentation, site content, or data across linked pages. CX Codex +1 80,700installs firecrawl-map Lists a site's URLs and filters them by a search term. Useful when you know the domain but not the exact page. CX Codex +1 80,200installs firecrawl-download Downloads an entire website or specific sections into local markdown files and screenshots. Useful for keeping offline docs ready for an agent. CX Codex +1 79,900installs firecrawl-interact Controls a live browser session on scraped web pages through text instructions or direct code. Useful when pages block simple scrapers with login walls or complex dynamic feeds. CX Codex +1 65,100installs firecrawl-monitor Watches page, site, or web-search changes and sends email or webhook alerts. Use it when a recurring check should replace repeated one-off scrapes. CX Codex +1 33,300installs

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor