Skip to main content

Privacy and security

Privacy policy

The local runtime is designed to work without a cloud account. Hosted cloud features process only the data needed for team sync, source-backed rules, optional embeddings, billing, and support.

Effective 2026-04-28 · Last updated 2026-04-28

Scope

This policy applies to cloud-connected CLI features, the hosted dashboard and APIs at https://difflore.dev, cloud sync, connected integrations, billing flows, support, and public product pages. Local-only CLI use remains on your machine unless you log in, sync, upload observations, or connect a cloud integration.

Data we collect

  • Account data: email, display name, avatar, OAuth provider IDs, team memberships, roles, and authentication/session records.
  • Product data: rules, candidate rules, rule metadata, origins, confidence signals, settings, review history, and related governance decisions.
  • Synced team-rules data: selected rules, candidate rules, source metadata, file patterns, evidence excerpts, review summaries, and integration metadata you explicitly send.
  • Limited code context: difflore does not upload full repository snapshots as a normal telemetry path. Cloud features process file paths, small evidence excerpts, source metadata, and accepted-outcome context only when you explicitly sync or request an opt-in workflow.
  • Billing data: billing customer IDs, subscription IDs, invoice references, plan status, and usage/overage records. The configured billing provider handles payment card numbers and payment-method details.
  • Operations data: IP address, user agent, request timestamps, endpoint usage, errors, webhook events, audit logs, and security logs.

How we use data

We use data to operate, secure, debug, and improve the service; sync rules and settings; provide retrieval and explainability workflows; run candidate governance, optional embeddings, demo previews, and team governance; prevent abuse; provide billing; and respond to support requests.

We do not sell personal data.

LLM providers and BYOK

The local-first runtime does not require DiffLore-hosted model calls for review mining or edit-time rule injection. Cloud model providers are used only for demo previews, optional embeddings, or explicitly enabled/contracted workflows such as BYOK.

Subprocessors

Depending on enabled features, deployment mode, and environment configuration, difflore may use these subprocessors or categories:

  • GitHub
  • Stripe
  • Anthropic
  • OpenAI
  • Voyage AI
  • RackNerd
  • Resend
  • Cloudflare

Processing regions currently configured for public disclosure: United States.

Retention and deletion

  • Local CLI data remains under your control on your machine.
  • Retained while your account or workspace is active, then deleted or anonymized according to the active retention policy.
  • Operational logs are currently retained for approximately 30 days unless needed longer for security, abuse prevention, legal, or billing records.
  • You may request export or deletion by contacting [email protected].

Security

difflore uses TLS in transit, access controls, audit logs, encrypted secrets where supported, scoped integration tokens, and hosted payment collection through Stripe. No system can guarantee absolute security.

DPA status: Available for Business and Enterprise plans during procurement.

SOC 2 status: Not certified yet; security documentation is available on request.

Your rights

Depending on your jurisdiction, you may have rights to access, correct, delete, export, or restrict processing of personal data. Contact [email protected] for privacy requests.

Contact