Meet Claudia

An open-source AI assistant who remembers across conversations, tracks your relationships, and helps you think strategically. Runs locally. No servers.

$npx get-claudia

Persistent Memory

Claudia remembers facts, preferences, and context across every conversation using a local SQLite database with vector embeddings.

Relationship Tracking

People are the primary organizing unit. Track communication style, history, what matters to them, and your current context together.

Commitment Tracking

Never drop a ball. Claudia tracks what you promised, what you’re waiting on, and surfaces deadlines before they hit.

Pattern Recognition

Notices recurring themes across conversations you might miss, like overcommitting or avoiding difficult decisions.

Gmail Integration

Search, read, and draft emails. Get morning inbox briefs. All through simple CLI commands with local-only token storage.

Google Calendar

View your schedule, search events, and create meetings through natural conversation. Claudia knows when you’re busy or free.

100% Open Source

Every line of code is public. No servers, no analytics, no tracking. Your data stays on your machine.

⭐ Apache 2.0 on GitHub

Terms of Service

Effective Date: March 2, 2026 · Last Updated: March 2, 2026

TL;DR: Claudia is free, open-source software under the Apache 2.0 license. It runs locally on your machine. There are no accounts, no subscriptions, no payments. Use it as you wish within the license terms.

1. What Claudia Is

Claudia is an open-source, locally-run AI assistant distributed as an npm package (get-claudia). She is designed to work alongside Claude Code as a memory and relationship layer.

Claudia runs entirely on your computer
There are no Claudia servers or cloud services
All data is stored locally in ~/.claudia/
No account creation is required

2. License

Claudia is licensed under the Apache License, Version 2.0. You may:

Use Claudia for personal or commercial purposes
Modify and distribute the software
Create derivative works

Subject to the conditions of the Apache 2.0 license, including attribution requirements.

3. No Warranty

Claudia is provided “AS IS” without warranty of any kind, express or implied, including but not limited to:

Warranties of merchantability
Fitness for a particular purpose
Non-infringement
Availability or reliability
Accuracy of any AI-generated content

4. Limitation of Liability

In no event shall the authors or copyright holders be liable for any claim, damages, or other liability, whether in an action of contract, tort, or otherwise, arising from, out of, or in connection with the software or the use or other dealings in the software. This includes but is not limited to:

Data loss or corruption
Security incidents
Inaccurate AI outputs or recommendations
Loss of business or revenue
Any indirect, incidental, or consequential damages

5. Your Data

You are solely responsible for your data. Claudia stores data locally on your machine in the ~/.claudia/ directory. We do not have access to, control over, or responsibility for your data.

Back up your data as you see fit
You can delete all Claudia data by removing the ~/.claudia/ directory
OAuth tokens for Google services are stored at ~/.claudia/tokens/

6. Third-Party Services

Claudia integrates with third-party services that have their own terms:

Google APIs (Gmail, Calendar): Subject to Google API Terms of Service
Ollama: Subject to Ollama’s license terms
Anthropic / Claude Code: Subject to Anthropic’s Terms of Service

Claudia is not responsible for the availability, accuracy, or terms of third-party services.

7. No Accounts or Payments

Claudia does not require account creation, subscriptions, or payments of any kind. There is nothing to cancel and no billing to manage.

8. AI-Generated Content

Claudia assists with generating text, extracting information, and recognizing patterns. All AI-generated content should be reviewed by you before use. Claudia does not provide legal, financial, medical, or professional advice.

9. Modifications

These terms may be updated from time to time. Changes will be posted in the GitHub repository. Continued use of Claudia after changes constitutes acceptance of the updated terms.

10. Governing Law

These terms shall be governed by and construed in accordance with the laws of the jurisdiction in which the developer resides, without regard to its conflict of law provisions.

11. Contact

For questions about these terms, open an issue on GitHub.

Privacy Policy

Effective Date: March 2, 2026 · Last Updated: March 2, 2026

TL;DR: Claudia runs entirely on your computer. Your data never leaves your machine. There are no servers, no analytics, no tracking. You can verify this yourself by reading the source code.

Data Storage

All data created by Claudia is stored locally on your computer:

DataLocationPurposeMemory database~/.claudia/memory/Memories, entities, relationshipsOAuth tokens~/.claudia/tokens/Google API access tokensConfiguration~/.claudia/config.jsonPreferences and settingsVault (optional)~/.claudia/vault/Obsidian-compatible markdown

No data is transmitted to the developer, to any server, or to any third party.

Google API Access

When you connect Gmail or Google Calendar via claudia google login:

What happens: Your browser opens to Google’s sign-in page. You grant access. OAuth tokens are saved locally.
What Claudia accesses: Only the data you explicitly request through CLI commands.

What Claudia does NOT do:

Store your Google data on any remote server
Share your data with anyone
Send analytics or telemetry
Access data beyond what you request
Run background syncs or polling

Google API Services User Data Policy

Claudia’s use of Google APIs complies with the Google API Services User Data Policy, including the Limited Use requirements:

Limited use: Google data is used only to provide the features you invoke.
No transfer: Google data is not transferred to third parties.
No advertising: Google data is never used for advertising.
No unauthorized access: Stored only on your local machine.

Revoking Access

You can revoke Claudia’s access at any time:

Run claudia google logout to remove local tokens
Visit Google Account Permissions to revoke the app

Ollama / AI Model Usage

Claudia uses Ollama (running locally) for text embeddings. All AI processing happens on your computer. No text is sent to external AI services by Claudia.

If you use Claudia within Claude Code, your conversations are subject to Anthropic’s privacy policy. Claudia’s memory system is separate and local.

Data Collection

Claudia collects zero data: no analytics, no telemetry, no crash reports, no usage statistics, no cookies, no tracking of any kind.

Children’s Privacy

Claudia does not knowingly collect data from children under 13. Since Claudia collects no data from anyone, this is inherently satisfied.

Open Source

Claudia is open source under the Apache 2.0 license. You can inspect every line of code:

Repository: github.com/kbanc85/claudia
OAuth implementation: cli/core/google-oauth.js
Token storage: ~/.claudia/tokens/

Disclaimer

Claudia is provided “as-is” without warranty of any kind. The developer assumes no liability for data loss, security incidents, or any other damages arising from use.

Changes to This Policy

Updates will be posted in the GitHub repository. Since Claudia has no server and collects no data, there is nothing to notify you about retroactively.

Contact

For questions, open an issue on GitHub.