# ketch

> A stateless command-line tool for web search, OSS code search, library docs, and scraping — one binary, no daemon, nothing to keep running.

For people, it replaces curl | pandoc or a browser tab. For agents, it gives predictable output, --json everywhere, and documented exit codes to branch on.

Install: curl -fsSL https://ketch.run/install | sh
The full manual, as Markdown: https://ketch.run/llms-full.txt

## Sections

- [Overview](https://ketch.run/#overview): Most research tooling for agents means wiring up several provider SDKs, each with its own auth and response shape.
- [Install](https://ketch.run/#install): To avoid piping a URL into a shell, read install.sh or download a prebuilt binary from the releases page.
- [Quickstart](https://ketch.run/#quickstart): No API key needed.
- [Choosing a command](https://ketch.run/#choosing-a-command): First match wins.
- [Commands](https://ketch.run/#commands): Thirteen commands.
- [Workflows](https://ketch.run/#workflows): Search, then fetch full content for every hit — bounded, because an unguarded page can cost ~25k tokens.
- [Research playbook](https://ketch.run/#research-playbook): These are the disciplines the bundled skill enforces, for agents and terminal users alike.
- [Backends](https://ketch.run/#backends)
- [Configuration](https://ketch.run/#configuration): Defaults live in ~/.config/ketch/config.json.
- [Exit status](https://ketch.run/#exit-status): Every failure mode has a number, so a script or an agent can branch on the outcome instead of pattern-matching an error string.
- [For agents](https://ketch.run/#for-agents): Instead of teaching an agent a search API, a code API, and a docs API — each with its own auth and response shape — give it one binary and five surfaces.
- [Notes](https://ketch.run/#notes)

## Links

- Install: https://ketch.run/#install
- Source: https://github.com/1broseidon/ketch
- Site as data: https://ketch.run/manifest.json
- Sitemap: https://ketch.run/sitemap.xml