-
Notifications
You must be signed in to change notification settings - Fork 1
Home
Turn live, localized search results into evidence-led research — then read the pages that matter.
SearchCans Agent Skills help AI agents run practical web-research workflows:
Search a real SERP → select useful sources → extract clean Markdown → deliver a traceable result.
Built for deep research, SEO/GEO planning, competitor analysis, RAG preparation, and webpage extraction audits.
Install all official SearchCans Skills:
npx skills add SearchCans/searchcans-skillsOr install the Deep Research workflow only:
npx skills add https://github.com/SearchCans/searchcans-skills --skill searchcans-deep-researchSet your API key in the execution environment:
export SEARCHCANS_API_KEY="your-api-key"PowerShell:
$env:SEARCHCANS_API_KEY = "your-api-key"Never place an API key in a prompt, source file, output artifact, or Git commit.
| Workflow | Best for | What the agent delivers |
|---|---|---|
| SearchCans Deep Research | Market, competitor, product, technology, policy, or company research | A bounded research bundle with a 3–5 question plan, extracted sources, uncertainty, and traceable URLs |
| SearchCans SERP Content Gap | SEO/GEO content planning, keyword research, search intent, and competitor-page analysis | A localized Google or Bing SERP evidence pack and content opportunity brief |
| SearchCans Reader SEO Audit | Web-to-Markdown extraction, RAG preparation, dynamic pages, and page-level SEO review | Extracted content plus canonical, H1, meta-description, and JSON-LD signals |
SearchCans combines three agent-ready building blocks under one API key:
- SERP API finds current, geo-targeted Google and Bing search evidence.
- Reader API converts selected web pages, PDFs, and Office documents into clean Markdown and optional HTML.
- Account API lets multi-request workflows check remaining credits and Parallel Lanes before work begins.
The Skills keep research disciplined:
- SERP snippets are leads, not proof.
- Reader-extracted URLs support consequential claims.
- Deep Research requires a 3–5 question research plan.
- Account-aware controls can cap scope and concurrency before a job exceeds its intended budget.
- Reports distinguish observed facts, inference, and uncertainty.
SearchCans Skills can use the Account API to make multi-request work more predictable:
- Estimate the credits required before a job starts.
- Stop or reduce a job when the available balance is insufficient.
- Limit simultaneous API calls to the account’s observed Parallel Lane count.
- Record only sanitized budget information — never account email, API keys, or tokens.
This is especially useful for multi-source research, multi-page SERP analysis, and higher-cost Reader proxy requests.
Sign in to SearchCans and visit Dashboard → Free Redemption Codes.
A new redemption code is released every week, and each code adds 1,000 API credits — enough for meaningful day-to-day testing and research.