Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Integrate Firecrawl `/scrape` into product code for single-page extraction. Use when an app already has a URL and needs markdown, HTML, links, screenshots, metadata, or structured page output. Prefer this skill over broader crawl patterns when the feature is page-level.
.claude/skills/sediman-agent-firecrawl-build-scrape/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -40% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 32% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -30% | 0% |
Use this when the application already has the URL and needs content from one page.
/interactmarkdown unless the feature truly needs another format.onlyMainContent for article-like pages where nav and chrome add noise./scrape as the default primitive for downstream LLM or indexing pipelines.Read the source-of-truth page for your project language before writing integration code:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 12,642 | 7,159 | -43% | 1 | 1 | 0% | 2,118 | 1,952 | -8% | 0 | 0 | — |
case-02 | pass→pass | 10,976 | 4,437 | -60% | 1 | 1 | 0% | 1,794 | 1,405 | -22% | 0 | 0 | — |
case-03 | pass→pass | 10,548 | 4,071 | -61% | 1 | 1 | 0% | 1,782 | 1,389 | -22% | 0 | 0 | — |
case-04 | pass→pass | 11,130 | 4,997 | -55% | 1 | 1 | 0% | 1,850 | 1,559 | -16% | 0 | 0 | — |
case-05 | pass→pass | 10,334 | 3,267 | -68% | 1 | 1 | 0% | 1,970 | 1,259 | -36% | 0 | 0 | — |
case-06 | pass→pass | 11,377 | 6,815 | -40% | 1 | 1 | 0% | 1,806 | 1,765 | -2% | 0 | 0 | — |
case-07 | fail→pass | 11,014 | 1,727 | -84% | 1 | 1 | 0% | 2,124 | 880 | -59% | 0 | 0 | — |
case-08 | fail→pass | 8,306 | 2,364 | -72% | 1 | 1 | 0% | 1,641 | 978 | -40% | 0 | 0 | — |
case-09 | fail→pass | 7,643 | 1,597 | -79% | 1 | 1 | 0% | 1,374 | 921 | -33% | 0 | 0 | — |
case-10 | fail→pass | 4,308 | 2,236 | -48% | 1 | 1 | 0% | 778 | 1,025 | +32% | 0 | 0 | — |
case-11 | fail→pass | 7,087 | 1,601 | -77% | 1 | 1 | 0% | 1,201 | 841 | -30% | 0 | 0 | — |
case-12 | fail→pass | 8,383 | 1,438 | -83% | 1 | 1 | 0% | 1,544 | 897 | -42% | 0 | 0 | — |
case-13 | fail→pass | 10,747 | 2,828 | -74% | 1 | 1 | 0% | 1,773 | 1,115 | -37% | 0 | 0 | — |
case-14 | pass→pass | 11,471 | 6,387 | -44% | 1 | 1 | 0% | 1,779 | 1,615 | -9% | 0 | 0 | — |
case-15 | pass→pass | 12,909 | 5,203 | -60% | 1 | 1 | 0% | 1,825 | 1,618 | -11% | 0 | 0 | — |
case-16 | pass→pass | 7,442 | 5,620 | -24% | 1 | 1 | 0% | 992 | 1,473 | +48% | 0 | 0 | — |
case-17 | pass→pass | 10,277 | 3,779 | -63% | 1 | 1 | 0% | 1,547 | 1,209 | -22% | 0 | 0 | — |
case-18 | pass→pass | 14,759 | 5,895 | -60% | 1 | 1 | 0% | 2,054 | 1,595 | -22% | 0 | 0 | — |
case-19 | pass→pass | 10,952 | 2,903 | -73% | 1 | 1 | 0% | 1,690 | 1,019 | -40% | 0 | 0 | — |
case-20 | fail→pass | 11,492 | 2,907 | -75% | 1 | 1 | 0% | 1,682 | 1,072 | -36% | 0 | 0 | — |
case-21 | pass→pass | 9,662 | 4,056 | -58% | 1 | 1 | 0% | 1,409 | 1,295 | -8% | 0 | 0 | — |
case-22 | pass→pass | 14,904 | 6,576 | -56% | 1 | 1 | 0% | 2,238 | 1,575 | -30% | 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. The headline lift of +36 percentage points is the difference between those two pass rates over the 22 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.
Other measured skills in the registry, with their headline benchmark lift.