OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Statewave

Statewave is an open-source, self-hosted memory runtime for AI agents that records events, compiles them into structured memories, and assembles reproducible, provenance-tagged context bundles.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Statewave — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Statewave when an agent needs structured memory across sessions and the application benefits from repeatable context bundles with source provenance.

Who it is for

  • Developers building AI agents that need information to persist across sessions.
  • Teams that want memory context traceable to the events it came from.
  • Application builders who want to run the memory layer alongside their own service.

Documented uses

  • Persist user preferences, decisions or history across agent sessions.
  • Compile recorded events into typed memories organized by a subject such as a user, account, agent or repository.
  • Retrieve ranked context bounded by a token limit for an application task.
  • Trace memories back to source episodes and inspect subject timelines.
Before you choose

Requirements & limits

The project is open source under the Apache-2.0 license. It is designed to run self-hosted; the repository documents Python and TypeScript SDKs and REST access.

  • Statewave is infrastructure to operate alongside an application, not a hosted memory service or chatbot framework.
  • The documented storage stack is Postgres with pgvector.
  • The default heuristic compiler is local and does not require an LLM API key; an LLM compiler and LiteLLM-based provider configuration are alternatives.
  • The API process is described as CPU-only; GPUs are only relevant if self-hosting an LLM compiler or embedding model.
  • Context assembly is token-bounded and described as deterministic for the same subject, task and point in time.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Use the Python quickstart memory loop

Record an event for a subject, compile its memories, and retrieve context for an application task.

Before you start

  • A running Statewave server reachable at its URL.
  • The Python Statewave SDK.
  • A subject identifier and event payload.

Steps

  1. Create a StatewaveClient pointed at the server URL.

  2. Call create_episode with a subject ID, source, type and payload.

  3. Call compile_memories for that subject.

  4. Call get_context with the subject, task and token limit, then use the assembled context in the application.

Expected outcome

The application receives ranked, token-bounded context assembled from the subject’s compiled memories.

Guide limitations

  • The example does not configure an LLM provider; the documented heuristic compiler can run without an LLM API key.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
12 October 2026
Original website title
smaramwbc/statewave · Official homepage title · retained in its source language
How we review sourcesSuggest a correction