Start
Quickstart
Install webctx and run your first search, URL read, and site map.
Install webctx
npm install -g webctx
webctx --version
Choose a task
Find documentation:
webctx search "Go context cancellation"
Search queries Brave, Tavily, and Exa when enabled and configured. You get one ranked list with short source snippets. It does not automatically open every result page.
Read a URL:
webctx read-link https://github.com/amxv/webctx
webctx read-link 'https://github.com/amxv/webctx/blob/main/README.md#L1-L20'
The first link returns a repository overview; the second asks for specific source lines. For normal web pages, webctx tries clean text and falls back to Firecrawl extraction.
Find pages on a site:
webctx map-site https://docs.firecrawl.dev
map-site returns a manageable page list, not an unbounded crawl of every URL.
Ask a focused question:
webctx read-link https://docs.example.com/api --question "How does pagination work?"
Question mode looks for relevant sections and can follow a small set of related documentation links.
Configure credentials
You only need keys for the capabilities you use:
| Task | Credentials |
|---|---|
| Search | One or more of BRAVE_API_KEY, TAVILY_API_KEY, EXA_API_KEY |
| Map sites or extract ordinary web pages | FIRECRAWL_API_KEY |
| Discover, inspect, or execute structured data | FIRECRAWL_API_KEY |
| Read most public GitHub URLs | Usually no token needed; GH_TOKEN helps with rate limits and permissions |
For example, put the keys you use into a .env.local file or your environment:
BRAVE_API_KEY=your-key
FIRECRAWL_API_KEY=your-key
Check what’s configured:
webctx status
See Credentials for Keychain setup and multiple accounts.
Work with structured data
Use three commands when you need data records rather than pages:
webctx research "podcast conversations about AI agents"
webctx inspect particle/podcasts/episodes/search
webctx execute particle/podcasts/episodes/search --inputs '{"semantic_search":"AI agents","limit":2}'
research and inspect discover available operations and their schemas. execute uses paid Firecrawl Alexandria credits. Read Research structured data before running paid queries.
Where next?
- Use webctx with agents for real workflows.
- Read a URL for GitHub and website examples.
- CLI reference for every flag.