search · davila7 — Claude Code skill: Google search via Bright Data
A skill that runs Google queries through the Bright Data SERP API and hands your agent structured JSON with links, titles and descriptions.
npx claude-code-templates@latest --skill web-data/searchWhat search does
The search skill gives Claude a way to query Google. A bash script, scripts/search.sh, calls the Bright Data SERP API and returns JSON with an organic array: link, title and a short page description. Your agent reads that JSON and keeps going — collecting sources, pulling facts, checking what actually ranks.
Two things have to exist before the first run: an API key from the Bright Data dashboard and an Unlocker zone created there. Both go into environment variables, BRIGHTDATA_API_KEY and BRIGHTDATA_UNLOCKER_ZONE. Without them the script does nothing, so check that first. Dependencies are curl for the request and jq for parsing; no jq on the machine means no parsed output.
You can also call it by hand: bash scripts/search.sh "query" [cursor]. The second argument is the page number, zero-indexed, default 0 — pass 1 for the second page.
What it helps with
- pull top Google results as structured data
- feed an agent links and snippets to work through
- page through search results
- build a source list for a topic without copy-paste
- wire web search into an agent through an external API
Example prompts for the agent
- run a search for "vector database benchmarks" and show me the links
- use the search skill and turn the results into a source list
- search for "rust async runtime" and grab the second page
How to install search
1claude-code-templates
npx claude-code-templates@latest --skill web-data/search2Skills CLI (Vercel)
npx skills add davila7/claude-code-templates --skill search -a claude-code3Manually: put the files in place
~/.claude/skills/search/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/search/
Manually: put the files in place
.agents/skills/search/SKILL.mdFor all projects use ~/.agents/skills/
Manually: put the files in place
.cursor/skills/search/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install davila7/claude-code-templates search2Manually: put the files in place
.github/skills/search/SKILL.mdPlatform CLI
gemini skills install https://github.com/davila7/claude-code-templates --path cli-tool/components/skills/web-data/searchManually: put the files in place
.devin/skills/search/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about search
How do I install search in Claude?
Use the templates installer: npx claude-code-templates@latest --skill web-data/search. Or npx skills add davila7/claude-code-templates --skill search -a claude-code. The skill lands at ~/.claude/skills/search/SKILL.md.
What does the search skill do?
It queries Google through the Bright Data SERP API and returns JSON with an organic array holding a link, title and description per result.
What does search need to run?
A Bright Data API key and an Unlocker zone, set as BRIGHTDATA_API_KEY and BRIGHTDATA_UNLOCKER_ZONE. It also needs curl and jq.
Similar skills
All skills for Browser and scraping →search by other authors comes first.
