> ## Documentation Index
> Fetch the complete documentation index at: https://docs.usealmanac.com/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> This handbook preview targets personal CLI 0.11.1. Check installed command help and the matching backend schema before using examples with another release. Examples with named people and placeholder IDs are fictional.

# Research the public web

> Choose ordinary page reads or a source's native search, then follow the returned evidence.

Almanac's public-web commands read information outside your saved knowledge. A search finds possible originals; a read retrieves one page or resource. Neither action saves a claim to your wiki. For a known URL, read it directly. For a question whose answer may be in a provider's collections, choose a source and its operation first.

```sh theme={"theme":{"light":"github-light","dark":"vesper"}}
almanac web search "Python asyncio documentation" --limit 3
almanac web read 'https://docs.python.org/3/library/asyncio.html'
```

The generic page reader is useful for ordinary pages and PDFs. A source such as Reddit or LinkedIn has native identities, filters and result shapes that a general URL read cannot supply. Use `almanac web sources --source reddit` to see its operations, then `almanac web describe reddit.post.search --engine netrows` for the exact contract. The [source directory](/web/sources) helps choose a source without reading the entire catalog.

Each command requests one upstream page or resource. Results may contain a next cursor, a resource to open, or an asset to download; none is followed automatically. A shipped operation, an available configuration and a successful live request are different facts. Check the returned notices and scope before calling an answer complete.

When a question depends on the original, keep its URL or native identity and read beyond the search snippet. [Web results](/web/results) explains saved output and follow-up resources. [Downloads](/web/downloads) explains why extracted PDF text, a video transcript and original file bytes are different evidence.