What the founder said
"These are the kind of tailwinds I want to lean into with the scribble works. If we are able to connect to agents that becomes an incredible distribution channel. Muse may make a lot of sense for us since it is consumer focused. Instagram is going to be a great way to sell this child education material too."
Source, in one line each
- Zuckerberg (09-18): Muse connectors open to developers. "You bring the API -- Muse brings the agent, the browser, and the context of what the person actually wants." Docs pointer: muse.ai/platform.
- Isenberg: connectors become the new app listings; your API matters more than your app; whoever you connect to becomes your landlord (points 3, 4, 12 of 13). Opinion, not reporting.
Ray's read
Fits the product exactly: parents on site, kids on paper, the printable is the deliverable, and the planner/customize rails are already API-shaped. The pitch line is "parent asks their agent for a Halloween hunt in Spanish for a 4-year-old; a PDF lands."
Ranked lanes: (1) Muse connector, consumer intent, largest surface if it hits scale; (2) Shopify Agentic Storefronts, same API, commerce side (already queued from the books wedge, see [[2026-09-02-commerce-agents-takeaways]]); (3) Instagram, distribution to humans, Reels of the printed thing in a kid's hands, real cohort photos only per [[DESIGN-scribble-works]].
Plan lives in the repo
RayDataCo/scribble-works#333 (filed 2026-09-19 12:54 ET under the plan-first rule (memory feedback_plan_in_gh_issues_before_build, founder 2026-09-19 10:23 ET)). Key facts from that research: no primary developer spec is published yet (muse.ai/platform is a submit-for-review page); the three engine-side gaps are agent-safe auth, a catalog read API (OpenAPI or a Model Context Protocol (MCP) endpoint), and an agent lane on the rate-limited generative rails. Those gaps are shared across every agent channel, which is why they come first.
Not decided
Whether to build any of it. Founder read of #333 pending.