For AI agents: a documentation index is available at /docs/llms.txt. Append .md to any page URL for markdown, or send Accept: text/markdown.
Command reference
amp groups its commands into a small set of surfaces. Discover them from the terminal:
bash
amp help # all product surfaces
amp help flags # commands within a surface
amp flags list --help # flags and examples for one command
amp version
amp auth status
Command families
auth: authenticate, manage profiles, and inspect credentials.context: resolve the active credential and org/project context.projects: list the projects your credential can access.events: manage taxonomy, listing, creating, updating, and deleting events.event-properties: manage the properties associated with an event.user-properties: manage a project's user properties.flags: manage feature flags, listing, creating, reading, updating, and archiving them.charts: list, inspect, and query saved charts.
Human tables vs. piped JSON
amp output adapts to who's reading it. In an interactive terminal, list and context commands print human-friendly tables and summaries. Pipe the output to a file or another command, or pass --json, and amp prints raw, compact API JSON instead, so scripts and agents don't spend effort parsing decoration.
Configuration documents --json and the other flags that apply across every command family, such as --dry-run and --yes.
Was this helpful?