websearch

name	websearch
description	Web search skill using DuckDuckGo (no API key needed) via the `ddgs` Python package. Use when the agent needs to discover URLs, search for information, or fetch and read web page content.

Websearch Skill

Search the web and fetch page content without any API key, using ddgs (DuckDuckGo Search) and lynx.

Requirements

pip install ddgs --break-system-packages   # already installed
which lynx                                  # for readable page content

Text Search

Returns title, URL, and snippet for each result.

python3 - <<'EOF'
from ddgs import DDGS

results = DDGS().text("your query here", max_results=5)
for r in results:
    print(f"Title : {r['title']}")
    print(f"URL   : {r['href']}")
    print(f"Snippet: {r['body'][:200]}")
    print()
EOF

News Search

python3 - <<'EOF'
from ddgs import DDGS

results = DDGS().news("your topic here", max_results=5)
for r in results:
    print(f"Title : {r['title']}")
    print(f"URL   : {r['url']}")
    print(f"Date  : {r['date']}")
    print(f"Source: {r['source']}")
    print()
EOF

Fetch a Page as Readable Text

Use lynx to strip HTML and get clean readable content:

lynx -dump -nolist "https://example.com" 2>/dev/null | head -100

For longer pages, pipe through grep or head/tail to focus on relevant sections:

lynx -dump -nolist "https://example.com" 2>/dev/null | grep -A5 "keyword"

Full Workflow: Search Then Read

python3 - <<'EOF'
from ddgs import DDGS
import subprocess

# 1. Search
results = DDGS().text("delta pager git syntax highlighting", max_results=3)
for r in results:
    print(r['title'], '->', r['href'])

# 2. Fetch the most relevant result
url = results[0]['href']
out = subprocess.run(
    ["lynx", "-dump", "-nolist", url],
    capture_output=True, text=True, timeout=15
)
print(out.stdout[:2000])
EOF

Gotchas

• DuckDuckGo may rate-limit heavy usage; add time.sleep(1) between requests if needed
• lynx renders static HTML only — JS-heavy SPAs may return partial content
• For JS-heavy pages, fall back to the browser skill (Playwright)
• Always respect max_results to avoid unnecessary traffic

When to Use

• Discovering URLs for a topic (curl/wget can't search)
• Getting up-to-date information not in training data
• Verifying current docs, changelogs, package versions
• News and recent events lookup