Minimal Example

The smallest valid PAM file — one memory entry with integrity hash:

{
  "schema": "portable-ai-memory",
  "schema_version": "1.0",
  "export_date": "2026-02-17T00:00:00Z",
  "owner": {
    "id": "user-123"
  },
  "memories": [
    {
      "id": "mem-001",
      "type": "skill",
      "content": "User is a cloud infrastructure engineer",
      "content_hash": "sha256:e1bae3ec291c99eced01fc91b4152a0cef541fccf2034fc11b3f90f4e4d79b6e",
      "temporal": {
        "created_at": "2026-02-15T00:00:00Z"
      },
      "provenance": {
        "platform": "chatgpt"
      }
    }
  ]
}

Field Annotations

• schema / schema_version — Identifies this as a PAM v1.0 file
• export_date — When this export was created (ISO 8601)
• owner.id — Unique identifier for the user
• memories[].id — Unique identifier for this memory
• memories[].type — Memory category (skill in this case)
• memories[].content — The actual memory content as a string
• memories[].content_hash — SHA-256 hash of the normalized content for integrity verification
• memories[].temporal.created_at — When the memory was first created
• memories[].provenance.platform — Which provider this memory originated from

Content Hash

The content_hash is computed from the normalized content string per spec §6:

1. Trim leading/trailing whitespace
2. Convert to lowercase
3. Apply Unicode NFC normalization
4. Collapse consecutive whitespace to single spaces
5. Compute SHA-256 hex digest

For "User is a cloud infrastructure engineer", after normalization (becomes "user is a cloud infrastructure engineer"), the hash is sha256:e1bae3ec291c99eced01fc91b4152a0cef541fccf2034fc11b3f90f4e4d79b6e.

Validate

Python

from jsonschema import Draft202012Validator
import json

with open("portable-ai-memory.schema.json") as f:
    schema = json.load(f)
with open("memory-store.json") as f:
    data = json.load(f)
errors = list(Draft202012Validator(schema).iter_errors(data))
print("valid" if not errors else f"{len(errors)} errors")

Node.js

import Ajv2020 from "ajv/dist/2020.js";
import {readFileSync} from "node:fs";

const ajv = new Ajv2020({allErrors: true});
const schema = JSON.parse(readFileSync("portable-ai-memory.schema.json", "utf8"));
const data = JSON.parse(readFileSync("memory-store.json", "utf8"));

const validate = ajv.compile(schema);
console.log(validate(data) ? "valid" : `${validate.errors.length} errors`);