Score templates

v1 hero score

• B2B SaaS Outbound

  Cold leads → reply triage → booked meetings

  Shipped

  Two agents working in concert: cold-leads finds candidates from Apollo, drafts personalized openers with Compose, queues them for review or auto-send via Gmail; reply-triage classifies inbound replies and advances the contact's stage. The Pipeline records every action.

  Skills used: apollo · compose · gmail · pipeline · notify Read the docs →

Planned

• Content cadence

  Editorial outline → drafts → handoff

  Planned

  Turn approved outlines into blog drafts on a cadence. Hands off to your editor agent (or a human) for review before publish.

• SEO performance

  Keyword rank tracking + alerts on drops

  Planned

  Watches your tracked keywords. Flags rank drops > N positions. Triggers a research agent to investigate the top SERP changes and propose a response.

• Customer support triage

  Inbox classification + routing

  Planned

  Reads incoming support mail, classifies (bug, billing, feature request, urgent), drafts a first-pass response for your review, and routes urgent issues into a notify channel.

Build your own — the Composer workspace

Power users will be able to author custom skills, agents, scores, and pipelines from inside Maestro itself in a dedicated Composer workspace — sibling to the catalog views, with in-browser skill editing, an agent builder, a score canvas, and inline test runners. v1 ships configure-and-run only (most operators want to tune what Maestro ships, not build new pieces); Composer lands in v1.5+ once we know what authoring patterns design partners actually want.

Until then, authoring happens by editing files in the repo and restarting the runtime. The Skills overview documents the Python SDK contract — every shipped skill is a working example.

Why so few templates?

Per the manifesto: a v1 with one hero score that genuinely works is more valuable than a v1 with thirty mediocre templates. Templates land when we've seen them work end-to-end on real workloads — not before. If you have a use case you'd like us to optimize for, apply for the open beta and tell us when you do.