---
title: "Memory overview"
description: "The four memory stores Cerevisor maintains, how they interact, what's local, and why memory matters."
slug: guides/memory-and-learning/overview
section: guides
subsection: memory-and-learning
canonical_url: https://cerevisor.com/docs/guides/memory-and-learning/overview
last_verified: 2026-05-18
last_verified_version: "1.2.0"
updated_at: 2026-05-18T15:08:18.053416+00:00
---

Cerevisor remembers things across sessions. Most LLM tools forget every conversation; Cerevisor builds up a persistent profile of you, your preferences, and what's worked in your past runs, and injects relevant context into every agent's prompt.

**All memory is local.** Stored under `~/.cerevisor/memory/`. No telemetry. No cloud sync. No backup that Cerevisor controls. If you want backup, you back up the folder.

## The four stores

### 1. User profile (`memory/user/`)

What you've told Cerevisor about yourself: your role, your goals, your preferences, your communication style. The harness occasionally proposes changes based on what it's learned; you approve or reject.

Editable directly from the Memory view → User tab.

### 2. Harness self-memory (`memory/harness/`)

What the harness has learned about its own runs, per-role metrics, learned tips, corrections, recent run logs. Auto-populated at end-of-run. Not editable.

The harness uses this to:

- Recommend better model choices ("you've used Sonnet on Researcher 12 times; Haiku produces the same quality at 10× lower cost").
- Surface recurring failure patterns ("Coach agent has been stalling on long inputs lately").
- Build retrospectives.

### 3. Meta-cognition (`memory/meta-cognition/`)

A layer above harness self-memory. After each run, the harness diffs current vs. prior self-memory snapshots and surfaces observations: "role X is drifting in success rate", "blind spot Y has persisted for 3 runs", "pattern Z has been reversing".

Includes a **harness self-portrait**: a compiled "what I've learned about working with you" summary that's refreshed on a schedule.

### 4. Freeform entries (`memory/entries/`)

Your own notes. Open the Memory view → Entries tab and type whatever you want. These can be referenced from agent instructions or just kept as personal scratch.

## How memory is injected

Every agent's system prompt includes a memory context blob. The blob is assembled at run time from:

- Your user profile (a short summary, not the whole thing).
- The harness's current self-portrait.
- Recent observations from meta-cognition (if relevant).
- Top-N freeform entries (ranked by recency + relevance).

The injection is capped to avoid blowing context budgets. Defaults are sensible; you don't need to tune them.

You can toggle injection on/off globally in **Settings → Memory** (the toggle is `metaCognitionInjectIntoPrompts`).

## Disagreements and proposals

When the harness wants to update your user profile (e.g. "user now prefers concise summaries based on recent edits"), it doesn't just rewrite; it surfaces a **proposal** in the Memory view → Proposals tab. You accept, reject, or edit before it's applied.

If a proposal conflicts with an existing user preference, it's flagged as a **disagreement** and queued under Disagreements until resolved.

## Versioning

Every monthly cycle, Cerevisor takes a snapshot of all memory docs and saves it under `memory/versions/`. You can browse old snapshots from Memory view → History.

Useful when you accidentally accepted a bad proposal or want to revert a memory change.

## What memory is NOT

- It's **not** a chat history. Cerevisor doesn't store conversation transcripts long-term. Audit logs cover that need.
- It's **not** crowdsourced. Your memory only contains what *your* runs produced. No data from other users.
- It's **not** shared. There's no marketplace, no aggregation, no egress. The architecture has hooks for future opt-in egress (see [Data inventory & consent](./data-inventory-and-consent.md)) but all egress flags default `false`.

## Memory health

The Memory view shows aggregate health: drift score, conflict count, staleness score. If something looks off (e.g. a profile that's contradicting itself), you'll see a banner in the canvas with a one-click "open Memory" link.

## Editing memory by hand

The memory files are markdown + JSON, openable in any editor. If Cerevisor's UI doesn't let you do something, you can edit the files directly. They live at `~/.cerevisor/memory/`. Restart Cerevisor after manual edits: the cache is refreshed on launch.