Shared AI Memory Starter Packet
Give your AI tools one maintained local memory, then run a visible intake automation with Claude or Codex. Bring in Markdown, RSS, YouTube, or Hacker News sources, browse source-tagged cards, and open the exact receipt for every Run.
One memory, one visible workflow
Connect the shared memory.
Use the onboarding prompt and audit-first setup tools to connect Codex, Claude Code, Hermes, or the tools you select to one maintained local source.
Run one intake source.
Choose Claude or Codex, start with the bundled Markdown sample, then add a Markdown, RSS, YouTube, or Hacker News source when you are ready.
Inspect every result.
Automation Home keeps source-tagged cards, historical Runs, and exact receipts visible. Keep it manual or add and disable an optional cadence without losing history.
What's inside
Shared AI memory vault and onboarding
One maintained local source for approved instructions, reusable skills, project routes, resources, system state, and a paste-ready onboarding prompt.
Cross-tool skill setup
The complete status-brief sample skill plus audit-first setup tools for Codex, Claude Code, Hermes, macOS, Linux, and Windows.
Automation Home with Fleet and Intake
A local interface for imported Claude history, packet automations, historical Runs, source-tagged cards, and exact record drawers.
Four ready-to-run source paths
Guided intake for Markdown, RSS or Atom, YouTube channels, and Hacker News, with stable source identity and original references.
Claude or Codex intake runner
Choose either executor, run the included sample or one real source, and keep the same card and receipt contract across both tools.
Receipts, safeguards, and optional cadence
Passed, failed, and no-op Run receipts; privacy and approval boundaries; plus a separately approved schedule with Run now and an off switch that preserves history.
Included with your membership
Enter your membership email and we'll send a download link that works for one year.