Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Discover URLs on a site. Use `--search` to find a specific page within a large site.
.claude/skills/firecrawl-map/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 84% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 125% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 164% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -48% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -54% | 0% |
Discover URLs on a site. Use --search to find a specific page within a large site.
Prerequisite: map requires authentication (no keyless free tier); without credentials the CLI prompts an interactive login.
bash# Find a specific page on a large site firecrawl map "<url>" --search "authentication" -o .firecrawl/filtered.txt # Get all URLs firecrawl map "<url>" --limit 500 --json -o .firecrawl/urls.json
Run firecrawl map --help for the full option list (sitemap handling, subdomains, etc.).
Done when: the URL list is saved under .firecrawl/ and you have selected the URLs to scrape or crawl next.
map --search to find the right URL, then scrape it.map https://docs.example.com --search "auth" → found /docs/api/authentication → scrape that URL.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 26,180 | 10,354 | -60% | 1 | 1 | 0% | 539 | 992 | +84% | 0 | 0 | — |
case-02 | fail→pass | 17,127 | 6,899 | -60% | 1 | 1 | 0% | 290 | 652 | +125% | 0 | 0 | — |
case-03 | fail→pass | 17,852 | 8,765 | -51% | 1 | 1 | 0% | 244 | 645 | +164% | 0 | 0 | — |
case-04 | pass→pass | 15,537 | 8,438 | -46% | 1 | 1 | 0% | 2,032 | 914 | -55% | 0 | 0 | — |
case-05 | fail→pass | 29,178 | 10,858 | -63% | 1 | 1 | 0% | 1,807 | 945 | -48% | 0 | 0 | — |
case-06 | fail→pass | 20,781 | 8,034 | -61% | 1 | 1 | 0% | 1,937 | 896 | -54% | 0 | 0 | — |
case-07 | fail→pass | 45,836 | 8,469 | -82% | 1 | 1 | 0% | 8,219 | 744 | -91% | 0 | 0 | — |
case-08 | fail→pass | 18,009 | 6,749 | -63% | 1 | 1 | 0% | 713 | 632 | -11% | 0 | 0 | — |
case-09 | fail→pass | 40,531 | 9,167 | -77% | 1 | 1 | 0% | 2,379 | 806 | -66% | 0 | 0 | — |
case-10 | fail→pass | 16,805 | 9,484 | -44% | 1 | 1 | 0% | 2,009 | 1,078 | -46% | 0 | 0 | — |
case-11 | fail→pass | 22,516 | 9,283 | -59% | 1 | 1 | 0% | 2,308 | 876 | -62% | 0 | 0 | — |
case-12 | fail→pass | 19,827 | 6,657 | -66% | 1 | 1 | 0% | 1,828 | 616 | -66% | 0 | 0 | — |
case-13 | pass→pass | 5,383 | 4,462 | -17% | 1 | 1 | 0% | 778 | 644 | -17% | 0 | 0 | — |
case-14 | fail→pass | 11,922 | 10,107 | -15% | 1 | 1 | 0% | 475 | 637 | +34% | 0 | 0 | — |
case-15 | fail→pass | 17,633 | 5,280 | -70% | 1 | 1 | 0% | 2,384 | 1,046 | -56% | 0 | 0 | — |
case-16 | fail→pass | 14,084 | 2,542 | -82% | 1 | 1 | 0% | 1,941 | 716 | -63% | 0 | 0 | — |
case-17 | fail→pass | 18,423 | 10,533 | -43% | 1 | 1 | 0% | 500 | 615 | +23% | 0 | 0 | — |
case-18 | fail→pass | 11,967 | 7,756 | -35% | 1 | 1 | 0% | 1,994 | 700 | -65% | 0 | 0 | — |
case-19 | pass→pass | 11,837 | 9,516 | -20% | 1 | 1 | 0% | 1,790 | 1,084 | -39% | 0 | 0 | — |
case-20 | pass→pass | 17,526 | 8,437 | -52% | 1 | 1 | 0% | 2,006 | 922 | -54% | 0 | 0 | — |
case-21 | pass→pass | 13,892 | 6,027 | -57% | 1 | 1 | 0% | 1,965 | 1,105 | -44% | 0 | 0 | — |
case-22 | pass→pass | 13,026 | 7,546 | -42% | 1 | 1 | 0% | 2,248 | 1,475 | -34% | 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 +73 percentage points is the difference between those two pass rates over the 17 comparable cases.
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 | +50% |
| gemini-3.6-flash | verified | 7/24/2026 | +32% |
Other measured skills in the registry, with their headline benchmark lift.