brave-headless
by kelexine
Web & Frontend DevelopmentBrief
Headless web search and content extraction via the Brave Search API.
Description
What This Skill Does
Headless web search and content extraction using the Brave Search API. Runs as a CLI tool without a browser, returning search results and full page content as plain text or JSON. Includes retry logic, circuit breaking, and bounded parallel fetching.
Combines search and content extraction in one tool with production-grade fault tolerance, avoiding the overhead of running a full browser like Playwright or Puppeteer.
When to Use It
- Searching the web from an AI agent workflow
- Extracting readable content from a URL for summarization
- Piping search results into downstream scripts
- Looking up documentation pages without a browser
- Researching topics and collecting structured output
Example Workflow
User asks: 'Find recent articles about WebAssembly and extract their content'
- 1
Runs search.js with the query and --content flag
- 2
Brave Search API returns top results with URLs
- 3
Content fetcher retrieves each page in parallel up to the concurrency limit
- 4
Readability extracts main article text, converted to Markdown
- 5
Results are printed as structured plain text or newline-delimited JSON
Numbered results with title, URL, snippet, and extracted Markdown content for each page
Requirements
- BRAVE_API_KEY environment variable
- Brave Search API account (free tier available at brave.com/search/api)
Installation
clawhub install kelexine/brave-headlessRun the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.