# microvms-agentd

> Run AI agents in sandboxed AWS MicroVMs from the CLI or Python, JavaScript, TypeScript, and Rust SDKs.

Read this first. It is the page written for a machine reader, and it names which surface answers
which question:

- [For agents](https://laithalsaadoon.github.io/microvms-agentd/agents.md)

The site has three tiers. Learn (`/learn/`) is task-shaped: the tutorials in order, then how-tos.
Reference (`/reference/`) is generated from `microvm manifest`, the binary's own statement of every
command, exit code, and response type, so a page there is the contract the binary ships. Internals
(`/internals/`) is the reasoning, in two kinds of document that are not equally reliable. The
hand-written documents — Platform, Protocol, Trust, Embedding, Strategy, Harness capabilities —
carry measured findings and design rationale, and they win any disagreement. The generated
categories — Architecture, Behavior, Analysis, Diagrams, Insights — were produced per-file from the
source tree, and every factual claim there carries a `path:line` citation pinned to the commit the
site was built from. Those citations anchor to line numbers, so they rot when the cited code moves
while still reading as authoritative.

Any page is available as Markdown: append `.md` to its path.

## Documentation Sets

- [Abridged documentation](https://laithalsaadoon.github.io/microvms-agentd/llms-small.txt): a compact version of the documentation for microvms-agentd, with non-essential content removed
- [Complete documentation](https://laithalsaadoon.github.io/microvms-agentd/llms-full.txt): the full documentation for microvms-agentd

## Notes

- The complete documentation includes all content from the official documentation
- The content is automatically generated from the same source as the official documentation