mcpField
G
google-serper-search

google-serper-search

github.com/shuliuzhenhua-sysProductivity0243 installsUpdated 32w ago
Repository

About google-serper-search

google-serper-search is a skill for running Google Serper searches through a Python script and presenting relevant results. It supports multiple search categories and geographic, language, and time filters.

Key features

  1. 01Run searches through the Serper API
  2. 02Search web, news, images, maps, and more
  3. 03Filter by country, language, and time
  4. 04Return results as JSON

Use cases

  1. 01Find recent news on a topic
  2. 02Search for images or places
  3. 03Narrow results to a country or language
  4. 04Retrieve structured results for summarization

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/shuliuzhenhua-sys/google-serper-search --skill google-serper-search

After installing, type/google-serper-searchin Claude Code to confirm it loaded.

Bundled files

References and scripts installed along with SKILL.md.

.gitignore
agents/
openai.yaml
references/
api_response.md
configuration.md
scripts/
serper_search.py

Details

From the project README

Google Serper Search

Use scripts/serper_search.py to call the Serper API and return JSON search results.

Check configuration

Confirm SERPER_API_KEY is available before searching.

Run:

printenv SERPER_API_KEY

If the variable is empty, read references/configuration.md and set it in the current shell or shell profile before continuing.

Run searches

Use:

python3 scripts/serper_search.py "query"

Use filters when the request needs them:

python3 scripts/serper_search.py "query" --type news --gl us --hl en --tbs "past week"

Supported --type values:

  • search
  • images
  • videos
  • places
  • maps
  • reviews
  • news
  • shopping
  • lens
  • scholar
  • patents
  • autocomplete

Aliases:

  • web -> search
  • image -> images
  • img -> images

Supported --tbs shortcuts:

  • past hour
  • past 24 hours
  • past week
  • past month
  • past year

Use --gl for country and --hl for language when the request needs regional or language-specific results.

Interpret results

Read references/api_response.md when you need the result schema.

For web and news results, prioritize:

  • knowledgeGraph
  • organic
  • peopleAlsoAsk
  • relatedSearches

For image results, prioritize:

  • image title
  • imageUrl
  • thumbnailUrl
  • dimensions
  • source page

Present the answer

Summarize the most relevant results instead of dumping raw JSON unless the user asks for the raw response.

Include source links in the final answer.

Call out missing configuration or API errors explicitly when the script returns an error field.

Comments

Markdown supported