curllm

curllm Examples

📚 Documentation Index

⬅️ Back to Main README

---

This document contains curated, end-to-end examples for common automation tasks with curllm. You can also generate runnable shell scripts for these examples into the examples/ directory.

• Generate scripts: make examples (or bash tools/generate_examples.sh)
• Scripts will be created as examples/curllm-*.sh
• Ensure services are running: curllm --start-services && curllm --status

Top 10 quick examples

1. Extract all links

   curllm "https://example.com" -d "extract all links"
   

2. Extract emails and phones

   curllm "https://example.com/contact" -d "extract all emails and phone numbers"
   

3. Take a screenshot

   curllm "https://example.com" -d "screenshot"
   

4. Products under 150 zł (public proxy rotation)

   export CURLLM_PUBLIC_PROXY_LIST="https://raw.githubusercontent.com/clarketm/proxy-list/master/proxy-list-raw.txt"
   curllm "https://ceneo.pl" -d "Find all products under 150zł and extract names, prices and urls" \
     --stealth --proxy rotate:public --csv -o products.csv
   

5. Products with registry rotation (after registering proxies)

   # register proxies via curlx or API, then:
   curllm "https://ceneo.pl" -d "Find all products under 150zł and extract names, prices and urls" \
     --stealth --proxy rotate:registry --html -o products.html
   

6. BQL: Hacker News links (CSS selectors)

   curllm --bql -d 'query { page(url: "https://news.ycombinator.com") { title links: select(css: "a.storylink, a.titlelink") { text url: attr(name: "href") } }}'
   

7. Visual fill contact form (stealth)

   curllm --visual --stealth "https://www.prototypowanie.pl/kontakt/" \
     -d "Fill contact form: name=John Doe, email=john@example.com, message=Hello"
   

8. Use a session (persist cookies between runs)

   curllm --session my-site "https://example.com" -d "screenshot"
   # later: same session reused automatically
   

9. Export results to XLS (Excel-compatible)

   curllm "https://example.com" -d "extract all links" --xls -o links.xls
   

10. WordPress: create a post using session

    curllm --session wp-s1 -d '{"wordpress_config":{"url":"https://example.wordpress.com","action":"create_post","title":"Hello","content":"Post body","status":"draft"}}'
    

Table of Contents

• Extract links from a page
• BQL: Hacker News links
• Fill contact form (visual + stealth)
• Login and download
• Visual scrape products
• CAPTCHA demo
• BQL JSON API
• Stealth scraping news titles
• Override model per-command
• POST with custom headers

---

LLM: Hacker News links

Command:

curllm "https://news.ycombinator.com" -d "Extract the page title and the first 30 news links. Use anchors matching CSS selectors 'a.titlelink' or 'a.storylink'. Return JSON shaped exactly as: {\"page\": {\"title\": string, \"links\": [{\"text\": string, \"url\": string}] } }" -v

Script:

• examples/curllm-llm-hn-links.sh

Notes:

• You can add --html -o hn.html to export links to an HTML table.
• For consistency with BQL example, the shape is {page: {title, links: [{text, url}]}}.

---

Extract links from a page

Command:

curllm "https://example.com" -d "extract all links" -v

Script:

• examples/curllm-extract-links.sh

---

BQL: Hacker News links

Command:

curllm --bql -d 'query {
  page(url: "https://news.ycombinator.com") {
    title
    links: select(css: "a.storylink, a.titlelink") { text url: attr(name: "href") }
  }
}' -v

Script:

• examples/curllm-bql-hn-links.sh

---

Fill contact form (visual + stealth)

Command:

curllm --stealth --visual \
  -d "Fill contact form: name=John Doe, email=john@example.com, message=Hello" \
  https://www.prototypowanie.pl/kontakt/ -v

Script:

• examples/curllm-fill-contact-form.sh

---

Login and download

Command:

curllm -X POST --visual --stealth \
  -d '{"instruction": "Login and download invoice", "credentials": {"user": "john@example.com", "pass": "secret"}}' \
  https://app.example.com -v

Script:

• examples/curllm-login-download.sh

---

Visual scrape products

Command:

curllm --visual "https://shop.com" -d "extract top 10 products with prices" -v

Script:

• examples/curllm-visual-scrape-products.sh

---

CAPTCHA demo

Command:

curllm --visual --captcha "https://example.com/captcha" -d "solve captcha and submit form" -v

Script:

• examples/curllm-captcha-demo.sh

Notes:

• For 2captcha, set CAPTCHA_API_KEY in .env. Otherwise local OCR will be attempted.

---

BQL JSON API

Command:

curllm --bql -d 'query { page(url: "https://example.com") { title links { text url } }}' -v

Script:

• examples/curllm-bql-json-api.sh

---

Stealth scraping news titles

Command:

curllm --stealth "https://news.ycombinator.com" -d "extract first 30 titles" -v

Script:

• examples/curllm-stealth-news.sh

---

Override model per-command

Command:

curllm --model qwen2.5:3b "https://example.com" -d "extract emails" -v

Script:

• examples/curllm-custom-model.sh

---

POST with custom headers

Command:

curllm -X POST -H 'Authorization: Bearer TOKEN' -H 'X-Trace: 1' \
  -d '{"instruction": "submit form with authenticated session"}' \
  https://httpbin.org/post -v

Script:

• examples/curllm-post-with-headers.sh

---

Export results (CSV/HTML/XML/XLS)

Use the CLI export flags to transform JSON results into tabular formats.

Requires: jq.

# CSV
curllm "https://ceneo.pl" -d "Find all products under 150zł and extract names, prices and urls" \
  --csv -o products.csv

# HTML table
curllm "https://example.com" -d "extract all links" --html -o links.html

# XML
curllm "https://example.com" -d "extract all emails" --xml -o emails.xml

# Excel-compatible (.xls generated as HTML table)
curllm "https://ceneo.pl" -d "Find all products under 150zł and extract names, prices and urls" \
  --xls -o products.xls

If -o is not provided, files are saved as curllm_export_YYYYMMDD-HHMMSS.(csv|html|xml|xls).

Tips

• Always start services first: curllm --start-services and verify with curllm --status.
• If ports change, .env is auto-updated. Use -v to see diagnostics.
• Visual tasks may use more GPU. Consider smaller models or reduced steps for testing.

This site is open source. Improve this page.