text_input stringclasses 2
values | intent_class stringclasses 2
values | recommended_node_properties dict | context_budget_allocation stringclasses 1
value |
|---|---|---|---|
Assignment: CubeSat Orbit Thermal Simulation Analysis due June 1st in ASE-401 | academic_deadline | {
"label": "CubeSat Orbit Thermal Simulation Analysis",
"type": "Note",
"tags": [
"canvas",
"assignment"
]
} | 400_chars |
Announcement: Lecture location changed to Engineering Lab 202 for Tuesday | academic_anomaly | {
"label": "Lecture Location Shift - Lab 202",
"type": "Context",
"tags": [
"canvas",
"announcement"
]
} | 400_chars |
Open Memory Standard (OMS) v1.0.0-beta
A Standardized Schema for Portable Personal AI Memories and Interconnected Context Nodes
1. Overview
The Open Memory Standard (OMS) is an open-source, vendor-neutral specification designed to eliminate data lock-in within generative AI models, contextual intelligence layers, and Personal Knowledge Management (PKM) platforms.
By standardizing personal memory payloads, applications can bridge the gap between unstructured multi-source context (such as academic dashboards, email schedules, and local markdown text) and vector databases without sacrificing user data ownership.
2. Platform Ingestion Objects
The baseline payload standard handles unstructured arrays exported by tier-one LLM providers and local scratchpads.
2.1 Explicit AI Memories (Platform Native Storage)
Captures explicit, atomic memory fragments stored as a flat array of unique string values.
{
"$schema": "[https://panderose.com/schemas/v1/explicit-memory.schema.json](https://panderose.com/schemas/v1/explicit-memory.schema.json)",
"provider": "openai",
"version": "1.0.0",
"extracted_at": "2026-05-24T12:22:00Z",
"memories": [
{
"memory": "User is a senior student working on a satellite thermal management design project."
},
{
"memory": "Prefers minimalist, infrastructure-grade, dual-color user interface layouts."
}
]
}
- Downloads last month
- 12