Archivst Agentic Usage

This hub is written for AI agents and other automation. Archivst.CLI is designed to be driven headlessly: every command runs once, exits, and reports results through three machine-readable channels:

• Exit codes — 0 = success, non-zero = error
• JSON notification files — one file per event, read then delete
• Structured log files — JSON lines and per-run summaries

Prefer passing values explicitly (--token, --code, --yes, --ftp-pass) over relying on interactive prompts, and use --quiet for minimal output. All pages are authored in Markdown; append ?raw=1 to any page URL (or load the linked file directly) to get the raw Markdown source.

Purpose

Archivst.CLI provides a cloud backup service to guard against unnesessarily uploading unchanging/infrequently-changed archival data to a secure cloud storage provider with the goal of affordably providing long term storage. In the case of compressible images, options to manage the compression when stored to the cloud are provided to further lower costs.

Topics

• Native CLI install — download and run the CLI binary on the local machine
• Preflight checks — verify the agent environment before any cloud operation
• Headless authentication — token sign-in, listener logins, and logout
• BackupSet JSON reference — file format, folder fields, and examples
• Notifications — machine-readable JSON event files
• Logs — log files and structured log queries
• Storage price estimates — estimate the monthly cost of storing a quantity of data via JSON
• Example agent workflow — a full backup/restore run from an agent's point of view
• cli-webui container — running Archivst in Docker for agent-driven setups

Raw Markdown

Every page above has a plain Markdown source at the same slug with a .md extension, for example /agentic/auth.md. This hub is served from /agentic/index.md.