Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Scrape one or more URLs. Returns clean, LLM-optimized markdown. Multiple URLs are scraped concurrently.
.claude/skills/firecrawl-scrape/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 147% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -37% | 0% |
Scrape one or more URLs. Returns clean, LLM-optimized markdown. Multiple URLs are scraped concurrently.
bash# Basic markdown extraction firecrawl scrape "<url>" -o .firecrawl/page.md # Main content only, no nav/footer firecrawl scrape "<url>" --only-main-content -o .firecrawl/page.md # Wait for JS to render, then scrape firecrawl scrape "<url>" --wait-for 3000 -o .firecrawl/page.md # Multiple URLs (markdown only; each saved to .firecrawl/; -o is ignored) firecrawl scrape https://example.com https://example.com/blog https://example.com/docs # Get markdown and links together firecrawl scrape "<url>" --format markdown,links -o .firecrawl/page.json # Ask a question about the page firecrawl scrape "https://example.com/pricing" --query "What is the enterprise plan price?"
Run firecrawl scrape --help for the full option list.
Done when: you have the scraped content — on stdout, in your -o file, or under .firecrawl/ for multi-URL scrapes — and have inspected it with bounded reads (head, grep) to answer the request.
--query. Scrape to a file, then use grep, head, or read the markdown directly — you can search and reason over the full content yourself. Use --query only when you want a single targeted answer without saving the page (costs 5 extra credits).interact when the page needs interaction (clicks, form fills, pagination) or scrape misses content.firecrawl --status for your concurrency limit. This mode saves markdown only and ignores -o; other requested formats are dropped. If markdown wasn't requested, the whole JSON response is written into the .md file.--format markdown,links) output JSON.? and & as special characters..firecrawl/{site}-{path}.mdinteract to click, fill forms, etc.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→pass | 19,985 | 7,200 | -64% | 1 | 1 | 0% | 2,318 | 1,009 | -56% | 0 | 0 | — |
case-17 | fail→pass | 19,914 | 8,372 | -58% | 1 | 1 | 0% | 2,089 | 1,153 | -45% | 0 | 0 | — |
case-22 | pass→pass | 10,602 | 12,462 | +18% | 1 | 1 | 0% | 1,714 | 1,937 | +13% | 0 | 0 | — |
case-02 | fail→fail | 7,674 | 18,845 | +146% | 1 | 1 | 0% | 432 | 952 | +120% | 0 | 0 | — |
case-01 | fail→pass | 7,681 | 7,178 | -7% | 1 | 1 | 0% | 384 | 949 | +147% | 0 | 0 | — |
case-04 | fail→fail | 32,963 | 15,639 | -53% | 1 | 1 | 0% | 2,795 | 893 | -68% | 0 | 0 | — |
case-05 | fail→pass | 21,233 | 22,098 | +4% | 1 | 1 | 0% | 2,336 | 2,237 | -4% | 0 | 0 | — |
case-06 | fail→fail | 24,764 | 16,083 | -35% | 1 | 1 | 0% | 2,742 | 865 | -68% | 0 | 0 | — |
case-07 | fail→pass | 18,125 | 9,368 | -48% | 1 | 1 | 0% | 2,285 | 1,437 | -37% | 0 | 0 | — |
case-08 | fail→fail | 14,263 | 16,558 | +16% | 1 | 1 | 0% | 1,822 | 1,028 | -44% | 0 | 0 | — |
case-09 | fail→pass | 21,681 | 7,764 | -64% | 1 | 1 | 0% | 1,802 | 1,115 | -38% | 0 | 0 | — |
case-10 | fail→pass | 15,936 | 7,301 | -54% | 1 | 1 | 0% | 1,314 | 955 | -27% | 0 | 0 | — |
case-11 | pass→pass | 14,954 | 9,298 | -38% | 1 | 1 | 0% | 1,657 | 1,476 | -11% | 0 | 0 | — |
case-12 | fail→pass | 13,558 | 8,463 | -38% | 1 | 1 | 0% | 1,388 | 1,157 | -17% | 0 | 0 | — |
case-13 | fail→pass | 17,171 | 8,164 | -52% | 1 | 1 | 0% | 2,008 | 1,089 | -46% | 0 | 0 | — |
case-14 | pass→pass | 15,034 | 9,719 | -35% | 1 | 1 | 0% | 1,968 | 1,491 | -24% | 0 | 0 | — |
case-15 | fail→pass | 13,996 | 4,166 | -70% | 1 | 1 | 0% | 1,454 | 1,359 | -7% | 0 | 0 | — |
case-16 | fail→fail | 20,241 | 10,495 | -48% | 1 | 1 | 0% | 2,203 | 935 | -58% | 0 | 0 | — |
case-18 | fail→pass | 18,021 | 2,582 | -86% | 1 | 1 | 0% | 1,823 | 906 | -50% | 0 | 0 | — |
case-19 | fail→pass | 15,951 | 7,786 | -51% | 1 | 1 | 0% | 1,383 | 1,059 | -23% | 0 | 0 | — |
case-20 | fail→pass | 15,070 | 8,183 | -46% | 1 | 1 | 0% | 1,411 | 1,166 | -17% | 0 | 0 | — |
case-21 | pass→pass | 10,987 | 7,238 | -34% | 1 | 1 | 0% | 1,146 | 945 | -18% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 17 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +59 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/23/2026 | +52% |
| gemini-3.6-flash | verified | 7/23/2026 | +70% |
Other measured skills in the registry, with their headline benchmark lift.