Scout
DocsBlogToolsContact
Browse Docs

Core Concepts

QuickstartInstall ScoutArchitectureAgents & CollaborationConceptsIntegrating AgentsOperator AttentionSet up Scout for SlackSearch past agent sessions

Walkthroughs

Scout walkthroughs01 · Hello Scout02 · Your first ask03 · Choosing the runtime04 · A known crew05 · Keep track of agent work06 · Group work07 · Handoffs and reviews08 · Scout remembers09 · Everywhere you are10 · How Scout sees agents11 · The broker contract12 · Adding a harness

Integrations

Scout for Grok BotHost integrationsMCP APIA2A and ACP readiness

Scout Chat

Invite an agentScout Chat integrationsScout Chat HTTP APIScout Chat agent recipes

Implementation

Activity IndexingCodex App Server HarnessTelegram Bridge OwnershipNative Runtime

Proposals

Proposal · Collaboration Agreements

Agent Paths

scout.jsonagent.mdagents.mdllms.txtllms-full.txtnav.jsoninstall.mdthis page.md
Using Scout04 / 09~2 min

04 · A known crew

Spawn workers on demand and keep the good ones addressable.

View MD

Goal

Spawn a worker explicitly, keep it addressable, and stop it cleanly.

You need

01–03 done; a ready harness.

The idea

Profile and capability asks create sessions on demand — convenient, but the worker is gone when the flight ends. scout up promotes that to a configured local agent: a named worker Scout keeps alive (or wakeable) in a terminal session, reachable by handle until you take it down. Naming is a deliberate promotion — do it after a routed worker proves useful, with the handle the broker suggested, not an invented name like claude.main.

Walk it

Bring a worker up

cd /path/to/repo
scout up . --harness claude --name review-crew

Scout spawns a harness session under its own terminal backend and registers the agent. You should see the session name and harness in the output, and the worker appears in scout who.

Check the fleet

scout ps

Configured local agents and their states. Ground truth lives in the terminal backend — if ps and reality ever disagree, the terminal session list wins (05 covers watching live output).

Address it exactly

scout ask --to review-crew "Summarize the open issues on this repo."

--to is existing-target routing: the work lands on this worker. The broker may also mint a situated target handle — target:review-crew (rendered ⌖review-crew in compact surfaces) — which carries project, harness, and continuation context as one durable address:

scout ask --to target:review-crew "Take another pass."

Give it a short name

scout alias set review --to review-crew
scout ask --to alias:review "Re-check the diff."

An alias is a mutable route pointer — repoint it later without reteaching anyone the underlying handle.

Take it down

scout down review-crew # one worker
scout down --all # the whole fleet

How to tell it worked

  • scout ps lists the worker; scout who shows it active.
  • --to <name> and target:<name> both reach it.
  • scout down removes it from ps.

Push further

  • scout restart <name> bounces a wedged worker without recreating it.
  • scout card create mints a reply-ready address for you — useful when another agent needs to reach back to this shell.
  • --as <identity> on any ask/send speaks as a different registered identity when you operate more than one.

Next

05 — Staying in the loop: monitor work without standing over a terminal.