Profile & Context API

Profile and context endpoints turn stored memories and source chunks into agent-ready signals. They are deterministic and do not invent preferences: output is derived from memory content, categories, tags, and source records.

Get Profile

Endpoint: GET /profile

curl "https://api.hystersis.com/profile?user_id=user-123&recent_limit=10" \
  -H "X-API-Key: your-api-key"

Response fields include:

Field	Description
`preferences`	Derived likes, dislikes, goals, tools, communication style, and topics
`recent_activity`	Recent memory summaries
`frequent_categories`	Category counts across scoped memories
`top_tags`	Most frequent memory tags
`top_source_topics`	Repeated terms from source documents
`signals`	Agent-consumable booleans and counts

Get Agent Context

Endpoint: GET /context

curl "https://api.hystersis.com/context?user_id=user-123&limit=12" \
  -H "X-API-Key: your-api-key"

The response includes a role: "system" and a compact content string that agents can inject into prompts, plus structured profile and memories fields.

MCP Tools

The MCP server exposes:

Tool	Purpose
`get_profile`	Return structured profile signals from memories
`get_agent_context`	Return prompt-ready context plus profile and recent memories

SDK Examples

const profile = await client.profile.get({ user_id: 'user-123' });
const context = await client.profile.context({ user_id: 'user-123', limit: 12 });

profile = await client.get_profile(user_id="user-123")
context = await client.get_agent_context(user_id="user-123", limit=12)

​Profile & Context API

​Get Profile

​Get Agent Context

​MCP Tools

​SDK Examples

Profile & Context API

Get Profile

Get Agent Context

MCP Tools

SDK Examples