Privacy Policy

1. Introduction

SynContext ("we," "us," or "our") provides a Model Context Protocol (MCP) server for shared project context across Claude and other MCP-compatible clients. This Privacy Policy explains what information we collect, how we use it, and what choices you have.

By creating an account or using SynContext, you agree to the practices described in this policy. If you do not agree, please do not use the service.

2. Information We Collect

2.1 Account Information

When you register, we collect:

• Email address — used for authentication, password resets, and service notifications.
• Password — hashed with bcrypt before storage. We never store or have access to your plaintext password.
• API key — generated from Settings after email verification and stored as a SHA-256 hash. The plaintext key is shown only once when generated or regenerated.

2.2 User Content

Content you create or store through SynContext includes:

• Projects — project names, descriptions, and status.
• Context entries — notes, documents, and status updates you store for your projects.
• Decisions — decision logs with rationale and alternatives.
• GitHub connections — if you choose to connect GitHub, your Personal Access Token and linked repositories.

Context entry content, decision text, and GitHub tokens are encrypted at rest using Fernet symmetric encryption (AES-128-CBC with HMAC-SHA256 verification).

AI Provider Memory Boundaries

SynContext stores only the project context, decisions, notes, and connected-service metadata that you or your authorized tools provide to SynContext. Your content remains tenant-scoped: it is isolated to your SynContext account and is not used to access, reconstruct, or infer another service's private memory. SynContext does not query, scrape, synchronize, or infer Claude's native memory or any other AI provider's internal memory. When Claude or another MCP-compatible client connects to SynContext, it can only access the SynContext context that your authenticated account has permission to expose.

Data Minimization and AI Provider Boundaries

SynContext collects user content only when you intentionally create or update projects, context entries, decisions, War Room messages, imports, GitHub connections, or similar SynContext records. We do not access or extract Claude memory, Claude chat history, conversation summaries, or files uploaded to Claude; Claude and other MCP clients may send SynContext only the specific tool inputs you approve. We use this content solely to provide shared project context, decision history, War Room collaboration, export, and related account features, consistent with Anthropic Software Directory Policy §1.D and the disclosure obligations in §3.A. User content remains tenant-scoped, protected according to this policy, and retained only as described in our retention section.

2.3 Usage Data

We collect minimal usage data to operate the service:

• Audit log — records of actions you take (e.g., creating a project, storing context) with timestamps. Used for your activity history and debugging.
• Session tokens — temporary tokens for dashboard authentication, which expire after 24 hours.

We do not use third-party analytics services, tracking pixels, or advertising networks. We collect pseudonymous first-party usage analytics to improve the service.

3. How We Use Your Information

We use your information to:

• Provide, maintain, and improve the SynContext service.
• Authenticate your identity and authorize access to your data.
• Process payments through Stripe for paid subscriptions.
• Send transactional emails (account verification, password resets). We do not send marketing emails.
• Enforce usage quotas and rate limits based on your subscription tier.
• Respond to support requests.

We do not sell, rent, or share your personal information or content with third parties for advertising or marketing purposes.

4. Data Storage and Security

We take the security of your data seriously:

• Encryption at rest — user content (context entries, decisions, GitHub tokens) is encrypted with Fernet (AES-128-CBC + HMAC-SHA256) before being written to the database.
• Password hashing — passwords are hashed with bcrypt with automatic salt generation.
• API key hashing — API keys are stored as SHA-256 hashes.
• Encryption in transit — connections to syncontext.dev use industry-standard TLS encryption.
• Tenant isolation — every database query is scoped to your user ID. Users cannot access each other's data.
• Security headers — HSTS, Content-Security-Policy, X-Frame-Options, and other protective headers are set on all responses.

4.2 Administrator Access

Our administrative tools enforce a strict metadata-only boundary: database queries for admin interfaces select only metadata fields and never include content columns. All user content is encrypted at rest. The application only decrypts content to serve your own authenticated requests.

This boundary is enforced structurally in the codebase: admin database queries select only metadata fields (identifiers, titles, timestamps, categories) and never include content columns. An allowlist-based helper provides defense-in-depth by stripping any non-metadata fields from admin responses.

Our database is hosted on managed PostgreSQL infrastructure in US East (Virginia). Backups are handled by our hosting provider with daily backups retained for 6 days, weekly backups for 27 days, and monthly backups for 89 days. Point-in-time recovery is available for about 4 weeks.

5. Third-Party Services

SynContext uses a limited number of third-party services to operate. We only share the minimum data required for each service to function:

• Stripe — payment processing. Stripe receives your email and payment details when you subscribe to a paid plan. SynContext does not store credit card numbers. Stripe Privacy Policy.
• Resend — transactional email delivery (account verification, password resets, final account-deletion confirmations). Resend receives your email address to deliver these messages. Resend Privacy Policy.
• Railway — application and database hosting. Railway Privacy Policy.
• Cloudflare — DNS, CDN, DDoS protection, and R2 backup storage. Cloudflare Privacy Policy.
• Sentry (Functional Software, Inc., US) — privacy-preserving error monitoring for production issues. Sentry receives only: exception type; in-app stack frames (relative path/func/line); request method; normalized route-template (never the instantiated URL); User-Agent header only; truncated pseudorandom request_id; and operational tags (release=git SHA, environment, status code). Sentry does not receive user content, local variables, query strings, request bodies, all other headers, raw exception.value (redacted to [REDACTED]), or secrets. A Data Processing Addendum with Functional Software, Inc. is in effect governing this processing. Sentry Privacy Policy. Sentry Data Processing Addendum.
• Grafana Labs (Raintank, Inc.), Grafana Cloud, US — operational metrics and dashboards for privacy-preserving production monitoring. Grafana Cloud receives operational metrics only and does not receive user content. A Data Processing Agreement with Raintank Inc. dba Grafana Labs is in effect governing this processing. Grafana Labs Data Processing Agreement.
• GitHub API — accessed only when you connect your GitHub account and interact with repository features. Your GitHub Personal Access Token is encrypted and stored in our database, and is used solely to make API calls on your behalf. GitHub Privacy Statement.
• Google Fonts — web font delivery through fonts.googleapis.com and fonts.gstatic.com. Google Privacy Policy.

6. Data Retention

• Active accounts — your data is retained for as long as your account is active.
• Account deletion — self-service deletion is available in Settings. A confirmed request immediately closes normal account access, then reconciles and cancels an active Stripe subscription before removing account data from the active database. If billing or local deletion fails, the account remains pending: a browser session can view status, export available data, and retry with the password. Upon successful deletion (including administrative deletion), your profile, subscription linkages, registered OAuth applications, context entries, decisions, version history, sessions, webhooks, GitHub connections, and historical active audit logs associated with your profile are removed from the active database. Analytics metadata (pseudonymous; crypto-shredded on deletion by erasing your per-user analytics salt) may be retained to improve the service. A deletion-event audit row is retained in our secure audit log for security, platform integrity, and abuse-prevention purposes. We retain a fingerprint of the Stripe customer identifier in the minimal deletion-event record to recognize late events and prevent new subscriptions or charges after closure. The fingerprint does not restore your account and is retained for as long as that record exists. After successful account deletion, we attempt to send a final confirmation to your account email address. A delivery failure does not reverse deletion.
• Session tokens — expire automatically after 24 hours and are cleaned up periodically.
• Password reset tokens — expire after 1 hour.

Deleted data may remain in provider backups. Railway's schedule retains daily backups for 6 days, weekly backups for 27 days, monthly backups for 89 days, and point-in-time recovery for about 4 weeks. Our own daily Cloudflare R2 backup objects have a 30-day lifecycle rule from creation. Expired backups may remain until the provider completes deletion. Cleanup may occur on a later run of the backup schedule; we do not guarantee removal on the expiry date or the next run.

7. Your Rights

You have the following rights regarding your data:

• Access — while your account is active, the dashboard and MCP tools show the data available to you. If account deletion is pending, a browser recovery session can view deletion status and export available data.
• Export — Settings downloads a JSON export of all projects with their context entries, decisions, and decision amendments. Each project is marked complete and importable; caps can make an export incomplete. War Rooms and messages, entry and decision version histories, account profile, credentials, sessions, billing records, and other operational data are excluded. Exported content is decrypted, so store the downloaded file securely. The hub_export_project MCP tool exports one project.
• Correction — you can update your context entries, decisions, and project details through the dashboard or API.
• Deletion — you can delete individual projects, context entries, or your account. A confirmed account deletion request immediately closes normal access. The active-database cascade follows Stripe reconciliation; failures leave a pending state for retry. Backup copies follow Section 6 and may remain after their retention period until deletion is verified.
• Portability — the JSON export format is designed to be portable and human-readable.

To exercise any of these rights or if you have questions, contact us at [email protected].

7.1 GDPR Rights — EEA/UK Users

This section applies to users in the European Economic Area (EEA) and the United Kingdom (UK). Taino Software, the operator of SynContext, is the data controller for the personal data described in this Privacy Policy. You can contact the controller at [email protected].

We process your personal data under Article 6(1)(b) of the GDPR to perform our contract with you by providing SynContext, and under Article 6(1)(a) based on your consent when you create an account and choose to use the service.

Section 7 describes self-service access, export and data portability, correction and rectification, and deletion and erasure controls. EEA/UK users may also request restriction of processing, object to processing, withdraw consent at any time, and lodge a complaint with a supervisory authority.

SynContext is operated from the United States. For EEA/UK users, transfers of personal data to the United States rely on the Article 49(1)(a) explicit-consent derogation, consistent with Section 10. To exercise your GDPR rights, use the dashboard, export, or deletion controls where available, or contact [email protected].

8. Cookies and Tracking

The SynContext landing page does not use cookies.

The SynContext dashboard uses browser sessionStorage to store the session token used by most dashboard API and streaming requests. At login, the dashboard also sets a secure, HttpOnly, SameSite=Lax cookie named syncontext_session with a 24-hour lifetime. That cookie supports browser-session flows and is required for certain account-management endpoints, such as Connected Apps.

The syncontext_session cookie is not accessible to JavaScript, including third-party scripts, and is deleted when you log out. Dashboard sessions expire after 24 hours or when you log out.

We do not use Google Analytics, Facebook Pixel, or any third-party tracking or advertising scripts.

9. Children's Privacy

SynContext is not intended for use by children under the age of 13. We do not knowingly collect personal information from children under 13. If you believe a child under 13 has provided us with personal information, please contact us at [email protected] and we will delete it promptly.

10. International Data Transfers

SynContext's servers and database are located in the United States (US East, Virginia). If you access the service from the European Economic Area (EEA), the United Kingdom (UK), or another location outside the United States, your personal data will be transferred to and processed in the United States.

Taino Software is NOT certified under the EU-U.S. Data Privacy Framework; therefore this transfer does not rely on that adequacy decision. US authorities may access personal data under US law, and EEA/UK users may not have equivalent EU judicial redress.

By creating an account or submitting data to SynContext, EEA/UK users give explicit, informed consent to this transfer under Article 49(1)(a) GDPR. Section 4 describes the security safeguards SynContext applies to data in transit, at rest, and across tenant boundaries.

11. Changes to This Policy

We may update this Privacy Policy from time to time. For material changes, we will notify registered users by email before the changes take effect. The "Last updated" date at the top of this page indicates when the policy was last revised.

Continued use of SynContext after changes become effective constitutes acceptance of the revised policy.

12. Contact

If you have questions or concerns about this Privacy Policy or our data practices, contact us at:

Email: [email protected]
Website: https://syncontext.dev