Context: Overview

​

Endpoint references

​

Choose an ingest payload

​

Typical call sequence

1. Ingest context: POST /context/ingest. The response returns an ID for each source.
2. Wait for indexing: poll GET /context/status with those IDs. A source is searchable at graph_creation and its graph is ready at completed. errored means indexing failed.
3. Retrieve context: POST /query.
4. Manage what is stored: list, inspect, update metadata, or delete by ID.

​

Key concepts

• Knowledge and memories: Knowledge is the documents, app sources, and facts your AI uses. Memories are preferences, decisions, and past conversations. Both live in a collection, and type: "all" on POST /query searches both from the same scope.
• IDs: /context/ingest returns an ID for each source. To assign your own, set id in the document_metadata, app_knowledge, or memories item. Use IDs to poll status, inspect content, and delete context.
• Metadata: Metadata Filters covers metadata fields, ingestion, updates, and filtering.
• Forceful relations: Links you declare between sources with relations.ids. In thinking mode, a query returns the linked sources in additional_context alongside the matching chunks. Set relations on the same document_metadata item as the file’s metadata:

import json

with open("runbook.pdf", "rb") as f:
    result = client.context.ingest(
        type="knowledge",
        database="acme_corp",
        documents=[("runbook.pdf", f, "application/pdf")],
        document_metadata=json.dumps([{
            "id": "runbook_deploy",
            "metadata": {"department": "ops"},
            "additional_metadata": {"owner": "platform-team"},
            "relations": {"ids": ["monitoring_guide"]},
        }]),
    )

​

Related

• Knowledge: forceful relations: linking sources at ingestion
• Context: memories and knowledge, and when to use which
• Metadata Filters: schema-backed and free-form metadata
• Query: Overview: retrieve ingested context