Skip to main content

SPID Protocol Specification

The SPID Protocol defines a structured, open standard for publishing, discovering, and interacting with Smart Packets via unique identifiers (PulseIDs).

This spec outlines how systems store, resolve, and route Smart Packets using the SPID identity and registry system.


1. Resolution Overview

Smart Packets are resolved by querying a SPID-compatible resolver using a PulseID or intent tag.

Input:

{
"query": "spid:creator:elena-podcast"
}

Output:
{
"packets": [
{
"id": "spid:creator:elena-podcast:intro",
"voice": "...",
"transcript": "...",
"intent": "intro/show",
"cta": [ ... ],
"meta": { ... }
}
]
}

The SPID Resolver MUST support:
Exact ID resolution (e.g. spid:brand:xyz-support)


Intent-based filtering (e.g. support/faq)


Tag- or metadata-based fallback



2. Identity Registry
Each PulseID maps to a JSON record that includes:
{
"id": "spid:brand:acme",
"displayName": "Acme Corp",
"inboxUrl": "https://api.acme.com/inbox",
"packetsUrl": "https://cdn.acme.com/spid/packets.json",
"metadata": {
"verified": true,
"created": "2025-05-28",
"tags": ["insurance", "mortgage", "quote"]
}
}

This allows systems to:
Fetch all published Smart Packets for a given identity


Route new voice messages to inboxes or bots


Determine if the source is verified



3. Trust & Security
SPID-compliant systems SHOULD:
Sign packets with a verifiable token or public key


Verify authorship before displaying packets to users


Respect expiration metadata (expires field)


Optional:
Use DNS-based TXT records to validate ownership of spid:domain:* IDs


Include HTTPS-based signatures for secure handoff between agents



4. Intent Structure
Intent strings follow a predictable path-based format:
[category]/[action] or [topic]/[context]

Examples:
insurance/quote


onboarding/welcome


product/demo


AI systems can use these intents for retrieval, clustering, or handoff logic.

5. Compliance Criteria
To be SPID-compliant, a system must:
Serve or retrieve packets matching the Smart Packet Format


Accept SPID PulseID queries via API or registry


Respect routing logic based on intent and packet metadata


Support voice + transcript pairing


Limit CTAs to defined schema types



Future Extensions
Planned extensions include:
Cross-agent packet routing


Distributed registry models (using DNS or DHT)


Live voice-to-Smart Packet encoding


Permissioned inboxes and async threads


Schema.org and JSON-LD integration for semantic indexing



The SPID Protocol turns knowledge into a retrievable, trusted, voice-first layer of the web — optimized for AI, but owned by people.

---