메뉴
Searches DuckDuckGo via its HTML endpoint using scrapling. Returns results as markdown (default), raw HTML, compact JSON, or YAML. Use when performing web searches from within an agent workflow without requiring a search API key or external service.
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
직접 명령은 검토 Prompt를 거치지 않습니다. 실행하기 전에 소스를 확인하세요.
npx skills add https://github.com/tangledgroup/tangled-skills --skill duckduckgo명령은 한 줄로 유지됩니다. 복사하기 전에 가로로 스크롤해 전체 내용을 확인하세요.
로컬 사본을 원하시나요? SkillsMP에서 현재 제공할 수 있는 파일을 다운로드하세요.
SOC 직업 분류 기준
SKILL.md 표시 중
| name | duckduckgo |
| description | Searches DuckDuckGo via its HTML endpoint using scrapling. Returns results as markdown (default), raw HTML, compact JSON, or YAML. Use when performing web searches from within an agent workflow without requiring a search API key or external service. |
Searches DuckDuckGo via its HTML endpoint (html.duckduckgo.com/html/) using scrapling for fetching and parsing. Supports four output formats — all sourced from the same HTML API:
.web-result HTML elements{title, url, domain, snippet} objectsNo API key or external search service required. Uses uvx 'scrapling[shell]' for ephemeral execution.
Run search.sh with a query — no format flag needed:
bash scripts/search.sh "tangled group"
Output is clean markdown produced by scrapling's --ai-targeted mode, streamed directly to stdout.
Request raw HTML when you need the full DOM structure:
bash scripts/search.sh --format html
Returns .web-result divs with all CSS classes, links, and snippets intact.
Get compact JSON for programmatic processing:
bash scripts/search.sh "tangled group" --format json
Output:
[{"title":"Tangled Group, Inc.","url":"https://tangledgroup.com/","domain":"tangledgroup.com","snippet":"We are a software development company..."},...]
Each result object contains only fields present in the HTML: title, url, domain, snippet. Missing fields are omitted.
Get structured YAML output:
bash scripts/search.sh "tangled group" --format yaml
Output:
- title: 'Tangled Group, Inc.'
url: 'https://tangledgroup.com/'
domain: 'tangledgroup.com'
snippet: |
We are a software development company...
All four formats use the same source — DuckDuckGo's HTML endpoint:
search.sh constructs the URL: https://html.duckduckgo.com/html/?q=<encoded-query>uvx 'scrapling[shell]' extract get fetches and isolates .web-result elements--ai-targeted flag produces clean, AI-friendly output--ai-targeted to preserve original DOM structureformat.py parses it using Python's built-in html.parserAll formats create temporary files in /tmp/ using mktemp with PID-based prefixes (ddg-search-${$}-XXXXXX) to ensure uniqueness under parallel execution. Files are automatically cleaned up on exit via a bash trap handler (covers normal exit, SIGINT, SIGTERM).
Query is passed to Python via stdin (not shell argument expansion) to prevent shell injection. urllib.parse.quote URL-encodes the query. Multi-word queries are handled correctly — "tangled group" becomes tangled%20group.
scrapling[shell])html.parser, json, sys, urllib.parse, re)