
Poppy 1.0 Launches With 100+ Companies Already on Board
Sierra Poppy 1.0: The Open Personal Agent Protocol for Agent-to-Business Commerce
TL;DR
Poppy 1.0, Sierra’s Personal Agent Protocol, establishes a universal standard so personal AI agents can discover, authenticate with, and take action inside any participating business. With more than 100 design partners confirmed by Sierra in 2026, including Shopify’s platform-wide merchant coverage and Meta’s Muse agent actively checking for the protocol on every site it visits, the launch arrives with real production weight behind it. Adoption starts with one discovery file and standard OAuth authorization controls.
Quick Takeaways
- Poppy 1.0 is Sierra’s open agent-to-business interoperability protocol, published in 2026 alongside a full implementation kit
- More than 100 design partners back the initial release; 60 new companies joined in the final phase before launch, per Sierra (2026)
- Shopify activated Poppy across its entire merchant base; Meta’s Muse agent proactively checks for a valid poppy.json on every website it visits, per Sierra (2026)
- Businesses implement Poppy by hosting a single JSON discovery file at /.well-known/poppy.json over HTTPS and configuring OAuth 2.0 authorization
- Poppy 1.0 layers on top of existing MCP and OpenAPI infrastructure rather than replacing either standard
What Is Sierra’s Poppy 1.0 Personal Agent Protocol?
Poppy 1.0 is Sierra’s Personal Agent Protocol: an open specification that lets any personal AI agent discover what a business offers, verify that business’s identity, and take authorized actions on a customer’s behalf through the company’s declared channels. Sierra published the full 1.0 specification at personalagentprotocol.org and released an implementation kit alongside it.
Poppy fills a gap in agentic commerce: personal AI agents can reason and act, but have no shared language for identifying which businesses are reachable, what they support, or how to authenticate safely. The fix is a minimal design built on conventions developers already know: a well-known URI for discovery and OAuth 2.0 for delegated authorization. Any team with existing HTTPS and OAuth infrastructure can implement Poppy without major re-architecture.
How Poppy 1.0 Connects Personal AI Agents and Businesses
Poppy 1.0 connects personal AI agents to businesses through a single JSON discovery file hosted at /.well-known/poppy.json over HTTPS. When a personal AI agent wants to interact with a business, it fetches that file first. Inside it, the company declares which interaction channels it offers (a business agent, a website, or a direct API), lists its supported sign-in options, and describes capability metadata covering what actions the agent can request.
That discovery-first design makes Poppy interoperable at scale. No bilateral registration is required; a conforming agent can reach a conforming business by resolving the well-known URL alone. Sierra modeled this on how email and the web handle addressability: publish one known location, and any compliant client finds it.
After discovery, the agent authenticates using OAuth 2.0, a mature delegated authorization framework (see also the Wikipedia overview of OAuth). OAuth lets the user grant the agent scoped permissions, monitor active sessions, and revoke access at any time, without the agent ever holding the user’s primary credentials. That boundary matters when agents act without constant human supervision.
Poppy 1.0 layers on top of existing infrastructure rather than replacing it. A business that already operates a Model Context Protocol server or an OpenAPI endpoint simply points its poppy.json at that existing channel. Poppy handles discovery and identity; the underlying transactional infrastructure stays unchanged.
Shopify, Meta Muse, and the 100-Plus Poppy Launch Partners
More than 100 companies are design partners for Poppy 1.0, with 60 new companies joining in the final phase before the specification was published, per Sierra (2026). Poppy’s value scales with participation, so hitting critical mass before launch was a core requirement.
Shopify activated Poppy across its entire merchant base, making every store immediately addressable by compliant personal AI agents. Because Shopify hosts a large share of independent e-commerce globally, that single decision extends Poppy’s reach across potentially millions of storefronts without requiring individual merchants to configure anything.
Meta’s Muse agent brings the demand side. Muse proactively checks for a valid poppy.json whenever it visits a website, automatically discovering Poppy-enabled businesses during ordinary browsing, per Sierra’s 2026 launch announcement. Every Muse-driven site visit becomes a potential agent-business interaction with no extra configuration from the site owner.
These two integrations model Poppy’s intended flywheel: more businesses publishing poppy.json increases the value of building agents that check for it, and more agents checking for it increases the incentive for businesses to publish. Shopify supplies business-side density; Muse supplies agent-side demand pressure.
How poppy.json Works: Discovery File Structure, Identity, and OAuth Authorization
The poppy.json file anchors every Poppy agent interaction through three structural concerns. Channel declarations tell an incoming agent whether to route its request through a dedicated business agent, a navigable website, or a direct API endpoint. Identity metadata lets agents confirm they are communicating with the legitimate business endpoint. Authorization configuration points agents toward the company’s OAuth server and specifies the permission scopes available for agent sessions.
OAuth governs the security boundary of every agent action within Poppy. Rather than storing or forwarding user credentials, the agent requests a scoped token from the company’s authorization server after the user grants explicit consent. Those tokens can carry narrow permissions and can be invalidated per session.
The Poppy 1.0 spec also defines how businesses can register with aggregated discovery services, letting agents search for Poppy-enabled businesses by category or capability before resolving a specific discovery file. That layer is marked as optional in the 1.0 release but points toward where the protocol is heading as adoption matures.
Did You Know?
Sierra designed Poppy’s well-known URI path, /.well-known/poppy.json, to follow the same IETF convention used by ACME challenge files for TLS certificate issuance. Any web server already configured for automated HTTPS certificate renewal can serve a Poppy discovery file through the same infrastructure path, with no additional routing setup required.
What Poppy 1.0 Means for Agentic Commerce and AI Automation
Poppy 1.0 replaces brittle custom integrations with a standard discovery and authorization layer. Today, most AI automation reaches businesses through a custom API integration built by a developer or a scraping-and-clicking approach that treats websites as human-only interfaces. Both break frequently and scale poorly across many target businesses, making them impractical for general-purpose personal agents.
When businesses publish a poppy.json, they offer agents a stable, company-maintained entry point rather than one that must be reverse-engineered. That stability helps everyone: the business gets more predictable agent traffic patterns, the agent developer skips brittle per-site connectors, and the end user sees more consistent task completion across different businesses.
For practitioners designing AI-driven customer journeys, Poppy means you can build workflows that assume a standard handshake between your company’s systems and your customers’ personal agents. Expose your capabilities through one discovery file and one authorization flow, and reach every agent that checks for Poppy support.
Poppy 1.0 also creates a new dimension of business visibility. A company that publishes a detailed, accurate poppy.json is more likely to be discovered and selected by personal agents operating on a user’s behalf. In an agentic commerce environment, discoverability becomes a protocol-compliance question alongside a search-ranking one, and the two do not necessarily overlap.
Did You Know?
Because poppy.json is a static file at a predictable URL, agent clients can cache it aggressively and check only for updates on subsequent visits. Teams managing high-traffic or multi-region deployments may find it more practical to serve and version the discovery file through a CDN edge configuration rather than a CMS publishing workflow, keeping the file update path separate from application deploys.
How to Implement Poppy 1.0: Publishing poppy.json and Configuring OAuth
Sierra released an implementation kit alongside the Poppy 1.0 specification covering three areas: schema guidance for constructing a valid poppy.json, OAuth configuration templates, and a validator that checks a live deployment against the full 1.0 spec.
Start by mapping your existing infrastructure to the channel types Poppy recognizes: a conversational agent, a structured website transaction flow, or an API with OpenAPI documentation. A valid poppy.json can declare a single channel; full three-channel coverage is not required for compliance.
Publishing the discovery file is straightforward for any team running HTTPS infrastructure. Create the JSON according to the specification’s schema, place it at /.well-known/poppy.json, and confirm the endpoint returns a 200 status with a JSON content type. Run the implementation kit’s validator to surface schema errors before any agent encounters the file in production.
OAuth configuration is the most complex step for businesses not already running OAuth 2.0. For platforms already on OAuth, Poppy requires defining agent-specific permission scopes and registering the authorization endpoint in your poppy.json. If starting from scratch, the OAuth 2.0 specification is the authoritative reference, and standard libraries exist for every major server runtime. After deployment, run representative agent workflows end-to-end; runtime issues such as CORS header mismatches, token lifetime errors, or missing capability metadata only surface under real request conditions.
Poppy 1.0 Limitations, Adoption Risks, and What Comes Next
Poppy 1.0 launches with more than 100 design partners, but most of the global business web is not yet Poppy-enabled, meaning agent discovery will return empty results for a large share of targets in the near term. Agent developers need graceful fallback strategies for businesses without a discovery file, whether defaulting to a direct API call, navigating the site conventionally, or notifying the user that agent access is unavailable.
Ecosystem fragmentation is a related concern. MCP is widely adopted as a model-to-tool protocol, and OpenAPI remains the most common machine-readable API description format in production. Poppy is designed to coexist with both, but businesses will weigh the overhead of maintaining multiple machine-readable contracts. The simplest mitigation is treating poppy.json as a lightweight index pointing to your existing MCP server or OpenAPI endpoint rather than replicating capability descriptions across files.
Security considerations apply primarily to OAuth scope design. Agents receiving broadly scoped tokens could take actions beyond what a user intended. Businesses should publish narrow, explicit permission scopes in their poppy.json and enforce those constraints server-side, not just in the discovery file. Consent language shown to users should name the specific actions being authorized, not just the name of the requesting agent.
The Poppy 1.0 spec intentionally defers aggregated discovery, richer capability semantics, and multi-step transactional session handling to later versions. A minimal 1.0 that teams can implement correctly is more valuable for ecosystem growth than a sprawling release with partial implementations across the design partner base.
Poppy 1.0 vs Related Protocols at a Glance
| Protocol | Primary Purpose | Discovery Mechanism | Auth Standard | Agent-Native |
|---|---|---|---|---|
| Poppy 1.0 | Agent-to-business interoperability | poppy.json at /.well-known/ | OAuth 2.0 | Yes |
| MCP | Model-to-tool connectivity | Server config or registry | Varies by server | Yes |
| OpenAPI | REST API description | openapi.json or Swagger UI | Varies by API | No |
| OAuth 2.0 | Delegated authorization | Authorization server metadata | OAuth 2.0 | No |
Sources: Sierra (2026), modelcontextprotocol.io, oauth.net/2, openapi.org. Poppy 1.0 is designed to layer on top of MCP and OpenAPI rather than replace either standard.
Poppy 1.0 Adoption Strategy: When and How to Deploy
Businesses that publish a Poppy discovery file early gain first-mover visibility to agents like Meta’s Muse that actively check for poppy.json on every site visit. If your company already supports a customer-facing API or a conversational agent channel, the marginal engineering cost of adding a poppy.json is low relative to the positioning advantage of appearing in Poppy-aware agent results from day one.
The more substantive tradeoff falls on the authorization side. If your platform handles user authorization through session cookies or proprietary token formats rather than OAuth 2.0, adding a compliant OAuth flow for agent access represents a real engineering investment. That question often resolves in favor of delegating OAuth to a dedicated identity provider rather than building the authorization server in-house.
Businesses with Shopify storefronts face the simplest adoption path: Shopify’s platform-level support means baseline Poppy coverage is already in place. The relevant decision for Shopify merchants shifts to which agent capabilities to expose at the Shopify API layer, and whether the default capabilities Shopify publishes on their behalf meet their needs or require a custom poppy.json to extend.
For teams building personal AI agents, Poppy 1.0 changes the integration strategy in a concrete way. Instead of maintaining bespoke connectors for each target business, you can build a single Poppy client that handles discovery, authentication, and channel routing generically, then apply it to every Poppy-enabled business. That is the same economic logic that made OAuth worthwhile for third-party app developers: one protocol implementation, many endpoints, no bilateral negotiation per site.
Conclusion
Sierra’s Poppy 1.0 addresses the missing infrastructure layer between personal AI agents and the businesses they need to reach. By standardizing discovery through a JSON file at a predictable URL and anchoring authorization in OAuth 2.0, the protocol gives both sides a shared contract to build against without bilateral agreements or custom integrations on a per-company basis.
The launch cohort, more than 100 design partners per Sierra (2026), including Shopify’s platform-wide coverage and Meta Muse’s active adoption, is substantial enough to treat Poppy as a production concern rather than a speculative standard. Teams building agentic workflows should review the full specification and assess where a discovery file fits their current infrastructure, then consult the Sierra announcement for the implementation kit and validator.
The protocol’s long-term success depends on the same forces that shaped OAuth and MCP: enough implementations on each side that skipping it becomes a competitive liability. That inflection point has not arrived yet, but 100 design partners and two major platforms are a meaningful first step toward it.
Frequently Asked Questions
- What is Poppy 1.0?
- Poppy 1.0 is Sierra’s Personal Agent Protocol, an open specification that enables personal AI agents to find participating businesses, verify their identities, and act on a user’s behalf through the business’s declared interaction channels, all without requiring custom bilateral integrations between each agent and each business.
- How does Poppy 1.0 work?
- Businesses host a JSON file at a fixed well-known URL. That file declares the interaction channels they offer, their OAuth authorization endpoint, and agent capability metadata. Any personal AI agent can fetch the file and begin an authorized interaction without prior registration between the two parties.
- How many companies support Poppy 1.0?
- According to Sierra (2026), more than 100 companies are design partners for the initial release, with 60 new companies joining in the final phase before the specification was published publicly.
- What do Shopify and Meta contribute to Poppy?
- Shopify activated protocol support for every store on its platform, extending Poppy’s commercial reach across potentially millions of storefronts in one decision. Meta’s Muse agent proactively looks for a valid discovery file during website visits, turning routine browsing sessions into potential agent-business interactions for any Poppy-enabled site.
- How can businesses adopt Poppy 1.0?
- Review the 1.0 specification and map your infrastructure to the channel types the protocol recognizes. Publish a valid poppy.json at /.well-known/poppy.json, configure OAuth permission scopes for agent access, and run the implementation kit validator to confirm your deployment before going live.