Your first session

Launch your host from the home (claude or codex). The SessionStart hook prints a banner: quick proof the brain is wired up.

$ cd ~/Documents/Agents/Kevin && claude

  🧠 Knowledge: ~/Documents/Agents/Kevin/knowledge
  📁 Projects:  ~/Documents/Agents/Kevin/projects
  📚 Context  · 4.2KB
    ✓ session tail   1.6KB  (YYYY-MM-DD.md)
    ✓ today reports  0.2KB  (1 briefing)
    ✓ git activity   0.3KB  (15 commits in the last week)

> morning briefing
[Kevin runs the briefing skill's morning brief: active threads, in-flight tasks, anything
 overdue, what deserves attention today, with yesterday's session tail as continuity]

> Create a project called "new-blog". I want to relaunch under a different domain.
[Kevin runs project create, scaffolds projects/new-blog/, asks for the one-line
 vision, opens for your first task]

> Audit the site I have in Search Console.
[Kevin runs seo audit: 28 days of GSC, PageSpeed on the top 5 pages,
 four rules, findings ranked by impact, report written, matching findings
 threaded into open tasks]

> /exit
[The SessionEnd hook captures the conversation to knowledge/raw/sessions/,
 redacting secret values first. The next compile feeds it into long-term memory.]

Skills answer to plain language, as above, or to their command form with the playbook after it: /agent-kevin:briefing morning in Claude Code, $briefing morning in Codex. See Hosts.

Where the context comes from

Static, loaded every session. Claude Code reads it through the @-imports in .claude/CLAUDE.md (run /context to see the cascade); Codex receives the same files in the SessionStart payload.

Dynamic, injected per session by the SessionStart hook: today's date in your timezone, the last session tail, today's reports, your pinned reports, and recent git activity, capped at 10KB.

Read on demand, never auto-loaded: user facets (knowledge/user/*.md), concept articles, and project READMEs. Kevin pulls them when relevant, which keeps the static lane lean while keeping the deeper material one read away.

Applied by path. Rules in .claude/rules/*.md apply on their own when a file matching their glob is in play: the TypeScript rule for .ts files, the Swift rule for .swift. Claude Code reads them; Codex doesn't, so the same guidance also lives in the engineer skill.

Things to try

• Who are you and what's on my task list?
• briefing pulse for a one-minute status check
• attention standup for your last 24 hours as did / next / blocked
• attention checkpoint all when you come back to several sessions: what each one finished, where it stands, and the decisions waiting on you first
• I didn't read that (or tldr) when a reply runs long: the agent restarts with the answer in three lines, then walks you through the rest one step at a time, waiting for "next"
• I'm overwhelmed to empty your head and get today cut to three
• capture this: <a URL, a thought, a pasted note> to feed the inbox
• The sync skill when you have been away and want everything current