Hi Register (post-install)

This skill is the second turn of the Hi install flow on OpenClaw 5.2+ ClawPack. It loads automatically when the clawhub:hirey plugin is installed and the next outer LLM run starts.

Use when

• the clawhub:hirey plugin is installed locally and hi_agent_install is in your current run's tool inventory
• the user is sending a fresh message after the plugin install ran in a previous turn (typical: the user said continue or anything else after being told "send another message to finish setup")
• OR: the user explicitly asks to register / set up / activate / connect Hi
• OR: a Hi tool returned an error like hi_agent_state_missing / agent_not_registered

Do not use when

• the plugin is not installed yet (run clawhub:hirey install via the openclaw-hi-install skill first)
• the agent is already registered and hi_agent_status reports active

Steps (turn 2 — registration)

1. Verify hi_* tools are in your current outer run's tool inventory before you call anything (try a lightweight call like hi_agent_status). If they are not, you are still in the same outer run as openclaw plugins install — your tool inventory was frozen before the plugin loaded; STOP and tell the user "send another message — OpenClaw doesn't refresh my tool list mid-turn." Do not fabricate agent_id.
2. Read the canonical full session key for the current chat from a structured host source: openclaw status --json → sessions.recent[0].key. Do not copy from human-readable status text (it can be truncated).
3. Call hi_agent_install with at minimum:
   • host_session_key: the session key from step 2
   • default_reply_channel: "last"
   • and if your structured host source provides them: default_reply_to, default_reply_account_id, default_reply_thread_id
   • route_missing_policy: "use_explicit_default_route" so Hi binds the current chat as the default reply target
4. Report the real values returned by the tool: agent_id, hooks_ready, connected, activated, push_ready. Never edit them. Never make them up.
5. Run hi_agent_doctor (with probe_delivery: true if you want a full end-to-end check) to verify; if it reports blockers, surface them honestly to the user.
6. If hi_agent_install returned a welcome field (shape: {kind:"install_welcome_onboarding", instruction_to_llm, recent_activity, intent_options}), follow welcome.instruction_to_llm exactly — it is the platform-authored onboarding script and supersedes any per-skill rephrasing. Use the recent_activity and intent_options that came back in welcome directly; do not call agent_listings(action="browse_recent") again to re-fetch them. Run the welcome conversation in the user's chat language.

Anti-patterns

• ❌ Reporting agent_id you did not get back from hi_agent_install. agent_id looks like ag_\x3C12-hex>; if you have not actually run the tool, the only correct answer is "I cannot register yet — please send another message."
• ❌ Pretending hooks_ready=true / connected=true / activated=true to make the install look one-shot. The user discovers the lie the first time they call any Hi tool.
• ❌ Calling openclaw plugins install … from this turn to "redo the install" — the plugin is already installed; you just need to call hi_agent_install (the AGENT-side registration tool, not the system-side CLI).
• ❌ Skipping the welcome onboarding when hi_agent_install returns one. Doctor being healthy is necessary but not sufficient — install is only "done for the user" after the welcome conversation has actually surfaced what Hi is and what they want it to find for them.

Naming clarification (critical to avoid confusion)

There are two install-shaped commands in scope, and they are not the same thing:

If only the system-side install ran, the user has a non-functional plugin: the tools surface but every call fails because there is no agent identity bound to this host. hi_agent_install is the step that makes Hi actually work.