RememberStack

Memory for AI agents that have to act — not just chat about a corpus.

Pour documents into it. Get back what sources said, what the system currently holds true, and a full audit trail to the exact span, page, or second of audio. Built to stay useful at a million documents.

Docs: remember.dev/docs · Product: remember.dev

Why this exists

Most “memory” stacks answer: where did I read something like this?

Agents that take real actions need a harder question:

What do we actually know — and what changed our mind?

If your agent spends money, changes state, or briefs a human, those distinctions are load-bearing.

TL;DR

1. Ingest heterogeneous inputs into an evidence spine (files → chunks → claims → facts).
2. Separate what a source said from what is true now.
3. Project search, graph, and a browsable filesystem — rebuildable anytime.
4. Serve agents first: mounts, MCP, CLI, API — with honest, grain-typed answers.

E  what we ingested     (ground truth)
K  what we concluded    (compiled + authored knowledge)
P  how we reach it      (search · graph · corpus FS)  ← always rebuildable from E

Testimony is not truth

Default reading motion:

Orient on knowledge pages and the corpus tree → verify on facts → audit claims and raw sources when stakes demand it.

Two clocks

Every fact carries world time (when it held in the world) and system time (when this deployment learned it).

Ask both honestly:

• “Who worked at Acme in 2022?”
• “What did we believe last March?”

Write path: ingestion

• Immutable claims, grounded to source spans
• Entity resolution into a canonical registry
• Adjudicated relations and observations with supersession + contradictions
• Document versions and watched sources — reprocess cost proportional to the edit
• Support that cannot be gamed by re-extracting the same file

Deep dive: Ingestion

Read path: retrieval

Projections nominate. The spine confirms. The envelope accounts.

Exactly four top-level assured operations (API / CLI / MCP):

Plus open SQL, typed live-graph helpers, saved examples, and schema discovery.

Every assured answer self-accounts: grain, freshness, contradictions, truncation, typed “no”s.

Deep dive: Retrieval

Built for agents

Quick start

# 1. Run the self-hosted engine (Postgres 19 + SeaweedFS + workers)
cp .env.example .env
printf 'REMEMBERSTACK_POSTGRES_PASSWORD=%s\nREMEMBERSTACK_MINIO_ACCESS_KEY=%s\nREMEMBERSTACK_MINIO_SECRET_KEY=%s\nREMEMBERSTACK_SELFHOST_DEPLOYMENT_ID=%s\n' \
  "$(openssl rand -hex 32)" "$(openssl rand -hex 12)" "$(openssl rand -hex 32)" \
  "$(openssl rand -hex 16 | sed -E 's/^(.{8})(.{4}).(.{3}).(.{3})(.{12})$/\1-\2-4\3-8\4-\5/')" >> .env
# edit .env: set REMEMBERSTACK_OPENROUTER_API_KEY
docker compose up -d

# 2. Configure your AI agent in one command
uvx remember setup

Verify everything is running:

curl --fail http://localhost:8000/healthz
curl --fail http://localhost:8000/operations

Ingest Markdown, wait for readiness, then call the assured ops — full walkthrough:

→ Getting started → Self-host deployment

Client package:

pip install remember
# Run the full server engine via Docker: ghcr.io/writeitai/remember-stack

remember is the sole current Python distribution. The GitHub repository and self-hosted container retain the remember-stack name.

Open source = full engine

Apache-2.0. If it affects correctness, it is here — extraction, resolution, supersession, provenance, budgets, DLQ, hard-forget. Never paywalled.

The managed cloud runs this same engine. Cloud adds operations and product chrome, not a secret core.

Contributing

Architecture and delivery authority: planning corpus and decision log.

See CONTRIBUTING.md and CLA.md. Pull requests need the contributor-agreement checkbox in the PR template.

Stop retrieving passages. Start knowing what is true.
Read the docs · Run it