Hindsight

Long-term memory for AI agents.

AI assistants forget everything between sessions. Hindsight fixes that with a memory system that handles temporal reasoning, entity connections, and personality-aware responses.

Why Hindsight?

• Temporal queries — "What did Alice do last spring?" requires more than vector search
• Entity connections — Knowing "Alice works at Google" + "Google is in Mountain View" = "Alice works in Mountain View"
• Agent opinions — Agents form and recall beliefs with confidence scores
• Personality — Big Five traits influence how agents process and respond to information

60-seconds step

1. Install the Hindsight All package (client + API)

pip install hindsight-all

2. Import your OpenAI API key

export OPENAI_API_KEY=xx

3. Run embedded server and client

import os
from hindsight import HindsightServer, HindsightClient

with HindsightServer(llm_provider="openai", llm_model="gpt-5.1-mini", llm_api_key=os.environ["OPENAI_API_KEY"]) as server:
    client = HindsightClient(base_url=server.url)

    client.put(agent_id="my-agent", content="Alice works at Google")
    client.put(agent_id="my-agent", content="Bob prefers Python over JavaScript")
    
    client.search(agent_id="my-agent", query="What does Alice do?")
    
    client.think(agent_id="my-agent", query="Tell me about Alice")

Documentation

Full documentation: hindsight-docs

• Architecture — How ingestion, storage, and retrieval work
• Python Client — Full API reference
• API Reference — REST API endpoints
• Personality — Big Five traits and opinion formation

License

MIT