Skip to main content

Persistent memory standard for AI agents — local, portable, zero config

Project description

rememb

Persistent memory for AI agents — local, portable, zero config.

rememb demo

AI agents (Windsurf, Cursor, Claude, Continue) forget everything between sessions.
rememb gives them a structured memory that lives in your project, belongs to you, and works with any agent.


The problem

Every developer using AI agents hits this wall:

Session 1: "We're using PostgreSQL, the auth module is at src/auth/, prefer async patterns."
Session 2: Agent starts from zero. You explain everything again.
Session 3: Same thing.

Existing solutions (Mem0, Zep, Letta) require servers, API keys, cloud accounts, and framework lock-in.
You just want the agent to remember your project.


The solution

.rememb/
  entries.json   ← structured memory (project, actions, systems, user, context)
  meta.json      ← project metadata

That's it. A JSON file in your project. Your agent reads it at the start of every session.


Install

pip install rememb

For semantic search support:

pip install rememb[semantic]

Quickstart

# Initialize in your project
rememb init

# Write memories
rememb write "Project uses FastAPI + PostgreSQL + async patterns" --section project
rememb write "User prefers direct answers, no filler text" --section user
rememb write "Auth module lives at src/auth/, JWT-based" --section systems

# Read everything (for the agent)
rememb read --agent

# Search semantically
rememb search "authentication"

# Get ready-to-use rules for your editor
rememb rules windsurf
rememb rules cursor
rememb rules claude
rememb rules continue
rememb rules vscode

Agent integration

Configure once. Works forever.

Run rememb rules <editor> to get the instructions for your editor, then paste them once. From that point on, your agent automatically reads and writes memory on every session.

rememb rules windsurf   # Windsurf / Cascade
rememb rules cursor     # Cursor
rememb rules claude     # Claude Code
rememb rules continue   # Continue.dev
rememb rules vscode     # VS Code + Copilot
Editor Where to paste
Windsurf / Cascade .windsurfrules at project root — or Settings → Cascade → Custom Instructions
Cursor .cursorrules at project root — or Settings → Rules for AI
Claude Code CLAUDE.md at project root (auto-read every session)
Continue.dev config.jsonmodels[].systemMessage
VS Code + Copilot .github/copilot-instructions.md at project root (auto-read by Copilot)

Memory sections

Section What to store
project Tech stack, architecture, goals
actions What was done, decisions made
systems Services, modules, integrations
requests User preferences, recurring asks
user Name, style, expertise, preferences
context Anything else relevant

Commands

rememb init              Initialize .rememb/ in current project
rememb write <text>      Write a memory entry (--section, --tags)
rememb read              Read all entries (--section, --agent, --raw)
rememb search <query>    Semantic search (falls back to keyword)
rememb import <folder>   Import .md/.txt/.pdf files into memory (--section, --recursive, --dry-run)
rememb rules [editor]    Print agent rules for windsurf/cursor/claude/continue/vscode

How search works

rememb search uses sentence-transformers for semantic similarity search locally.
No API calls. No embeddings sent to the cloud. Falls back to keyword search if the model isn't available.


Design principles

  • Local first — everything is a JSON file in your project
  • Portable — copy .rememb/ and it works anywhere
  • Agnostic — works with any agent that can run CLI commands
  • Zero configpip install rememb && rememb init and you're done
  • No lock-in — plain JSON, read it with anything

Roadmap

Coming soon

  • rememb import <folder> — import .md, .txt, .pdf files into memory (--section, --recursive, --dry-run)
  • rememb delete <id> — remove a memory entry
  • rememb edit <id> — update an existing entry
  • rememb export — export memory to Markdown / Obsidian / Notion

Planned

  • MCP server (rememb mcp) — native IDE integration, no CLI required
  • rememb sync — sync ~/.rememb/ across machines via private git
  • rememb web — local browser UI to manage memories visually
  • VS Code / Windsurf extension

Contributing

git clone https://github.com/LuizEduPP/rememb
cd rememb
pip install -e ".[dev]"

PRs welcome. Issues welcome. Stars welcome. 🌟


License

MIT

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

rememb-0.1.4.tar.gz (1.1 MB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

rememb-0.1.4-py3-none-any.whl (10.8 kB view details)

Uploaded Python 3

File details

Details for the file rememb-0.1.4.tar.gz.

File metadata

  • Download URL: rememb-0.1.4.tar.gz
  • Upload date:
  • Size: 1.1 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.3

File hashes

Hashes for rememb-0.1.4.tar.gz
Algorithm Hash digest
SHA256 c20bf60084e88206c54000cb028ccfd762965550925aad488b8389d188f9fb0a
MD5 bf2c65d97006b118548f70f5e9f0f7f7
BLAKE2b-256 6be981e5febdd5070da1272f7d56e17464fd2705c7854b2ff8f5c0f1da618780

See more details on using hashes here.

File details

Details for the file rememb-0.1.4-py3-none-any.whl.

File metadata

  • Download URL: rememb-0.1.4-py3-none-any.whl
  • Upload date:
  • Size: 10.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.3

File hashes

Hashes for rememb-0.1.4-py3-none-any.whl
Algorithm Hash digest
SHA256 94f09e4672165798754f82126ab1ce73072e6928fb919393ddd7b28a21b2806c
MD5 67e6cbeb9c953ffe792ad2b4adc43e65
BLAKE2b-256 845cf3980355277052096986e7ad6b3ae6a37384058709f4f108f8f6b4573f40

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page