Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Fetch any URL and convert to markdown using Chrome CDP. Supports two modes - auto-capture on page load, or wait for user signal (for pages requiring login). Use when user wants to save a webpage as markdown.
.claude/skills/freestylefly-canghe-url-to-markdown/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 194% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 13% | 0% |
Fetches any URL via Chrome CDP and converts HTML to clean markdown.
Important: All scripts are located in the scripts/ subdirectory of this skill.
Agent Execution Instructions:
SKILL_DIR${SKILL_DIR}/scripts/<script-name>.ts${SKILL_DIR} in this document with the actual pathScript Reference: | Script | Purpose | |--------|---------| | scripts/main.ts | CLI entry point for URL fetching |
Use Bash to check EXTEND.md existence (priority order):
bash# Check project-level first test -f .canghe-skills/canghe-url-to-markdown/EXTEND.md && echo "project" # Then user-level (cross-platform: $HOME works on macOS/Linux/WSL) test -f "$HOME/.canghe-skills/canghe-url-to-markdown/EXTEND.md" && echo "user"
┌────────────────────────────────────────────────────────┬───────────────────┐ │ Path │ Location │ ├────────────────────────────────────────────────────────┼───────────────────┤ │ .canghe-skills/canghe-url-to-markdown/EXTEND.md │ Project directory │ ├────────────────────────────────────────────────────────┼───────────────────┤ │ $HOME/.canghe-skills/canghe-url-to-markdown/EXTEND.md │ User home │ └────────────────────────────────────────────────────────┴───────────────────┘
┌───────────┬───────────────────────────────────────────────────────────────────────────┐ │ Result │ Action │ ├───────────┼───────────────────────────────────────────────────────────────────────────┤ │ Found │ Read, parse, apply settings │ ├───────────┼───────────────────────────────────────────────────────────────────────────┤ │ Not found │ Use defaults │ └───────────┴───────────────────────────────────────────────────────────────────────────┘
EXTEND.md Supports: Default output directory | Default capture mode | Timeout settings
bash# Auto mode (default) - capture when page loads npx -y bun ${SKILL_DIR}/scripts/main.ts <url> # Wait mode - wait for user signal before capture npx -y bun ${SKILL_DIR}/scripts/main.ts <url> --wait # Save to specific file npx -y bun ${SKILL_DIR}/scripts/main.ts <url> -o output.md
| Option | Description | |--------|-------------| | <url> | URL to fetch | | -o <path> | Output file path (default: auto-generated) | | --wait | Wait for user signal before capturing | | --timeout <ms> | Page load timeout (default: 30000) |
| Mode | Behavior | Use When | |------|----------|----------| | Auto (default) | Capture on network idle | Public pages, static content | | Wait (--wait) | User signals when ready | Login-required, lazy loading, paywalls |
Wait mode workflow:
--wait → script outputs "Press Enter when ready"YAML front matter with url, title, description, author, published, captured_at fields, followed by converted markdown content.
url-to-markdown/<domain>/<slug>.md<slug>: From page title or URL path (kebab-case, 2-6 words)<slug>-YYYYMMDD-HHMMSS.md| Variable | Description | |----------|-------------| | URL_CHROME_PATH | Custom Chrome executable path | | URL_DATA_DIR | Custom data directory | | URL_CHROME_PROFILE_DIR | Custom Chrome profile directory |
Troubleshooting: Chrome not found → set URL_CHROME_PATH. Timeout → increase --timeout. Complex pages → try --wait mode.
Custom configurations via EXTEND.md. See Preferences section for paths and supported options.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→fail | 4,618 | 6,199 | +34% | 1 | 1 | 0% | 713 | 1,313 | +84% | 0 | 0 | — |
case-09 | fail→pass | 9,642 | 1,988 | -79% | 1 | 1 | 0% | 1,563 | 1,317 | -16% | 0 | 0 | — |
case-01 | fail→fail | 11,421 | 6,646 | -42% | 1 | 1 | 0% | 1,976 | 1,341 | -32% | 0 | 0 | — |
case-02 | fail→fail | 4,294 | 6,724 | +57% | 1 | 1 | 0% | 619 | 1,380 | +123% | 0 | 0 | — |
case-04 | pass→fail | 7,341 | 3,340 | -55% | 1 | 1 | 0% | 1,206 | 1,312 | +9% | 0 | 0 | — |
case-05 | pass→pass | 4,867 | 3,476 | -29% | 1 | 1 | 0% | 889 | 1,630 | +83% | 0 | 0 | — |
case-06 | fail→pass | 7,267 | 12,437 | +71% | 1 | 1 | 0% | 1,112 | 3,265 | +194% | 0 | 0 | — |
case-07 | fail→pass | 11,487 | 2,947 | -74% | 1 | 1 | 0% | 1,779 | 1,474 | -17% | 0 | 0 | — |
case-08 | fail→pass | 4,971 | 1,834 | -63% | 1 | 1 | 0% | 872 | 1,327 | +52% | 0 | 0 | — |
case-10 | fail→pass | 8,081 | 2,430 | -70% | 1 | 1 | 0% | 1,189 | 1,344 | +13% | 0 | 0 | — |
case-11 | pass→pass | 8,148 | 2,227 | -73% | 1 | 1 | 0% | 1,186 | 1,374 | +16% | 0 | 0 | — |
case-12 | fail→pass | 10,618 | 2,589 | -76% | 1 | 1 | 0% | 1,681 | 1,438 | -14% | 0 | 0 | — |
case-13 | fail→pass | 14,471 | 2,754 | -81% | 1 | 1 | 0% | 2,372 | 1,471 | -38% | 0 | 0 | — |
case-14 | fail→pass | 12,016 | 2,208 | -82% | 1 | 1 | 0% | 1,908 | 1,452 | -24% | 0 | 0 | — |
case-15 | fail→pass | 7,945 | 1,425 | -82% | 1 | 1 | 0% | 1,296 | 1,185 | -9% | 0 | 0 | — |
case-16 | fail→pass | 5,567 | 1,604 | -71% | 1 | 1 | 0% | 751 | 1,205 | +60% | 0 | 0 | — |
case-17 | fail→pass | 9,430 | 2,529 | -73% | 1 | 1 | 0% | 1,567 | 1,439 | -8% | 0 | 0 | — |
case-18 | fail→pass | 14,706 | 3,916 | -73% | 1 | 1 | 0% | 2,084 | 1,652 | -21% | 0 | 0 | — |
case-19 | pass→pass | 11,126 | 3,878 | -65% | 1 | 1 | 0% | 1,562 | 1,522 | -3% | 0 | 0 | — |
case-20 | pass→pass | 9,473 | 1,855 | -80% | 1 | 1 | 0% | 1,326 | 1,269 | -4% | 0 | 0 | — |
case-21 | fail→pass | 5,701 | 1,836 | -68% | 1 | 1 | 0% | 938 | 1,268 | +35% | 0 | 0 | — |
case-22 | fail→pass | 9,831 | 2,363 | -76% | 1 | 1 | 0% | 1,326 | 1,348 | +2% | 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 19 counted toward the lift figure. The other 3 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 19 comparable cases. 1 case got worse with the skill loaded, and it is 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.
Other measured skills in the registry, with their headline benchmark lift.