Privacy Policy

Last updated: November 9, 2026

Draft for review. Bracketed fields must be completed, and the text reviewed by counsel, before publication.

This Privacy Policy explains how [LEGAL ENTITY NAME] ("Midas," "we," "us," or "our") collects, uses, shares, and retains personal information when you use the Midas coding agent, the command-line interface, the dashboard at app.trymidas.io, the documentation at trymidas.io, cloud sandboxes, and GitHub pull request reviews (together, the "Service").

We are the controller of the personal information described here, except where a customer organization acts as the controller of its members' information. In that case, we process that information on the organization's behalf as a processor, and the organization's own privacy notice applies.

1. Information we collect

Information from sign-in

We do not issue passwords. You sign in with Google, GitHub, or a single sign-on provider configured by your organization. We receive:

  • Your name, email address, and profile image URL, where the provider supplies them.
  • The identity provider you used, your provider account identifier, and whether the provider has verified your email address.
  • For GitHub sign-in, an OAuth access token with the scopes read:user, user:email, and read:org. The token is stored in your encrypted session cookie and used server-side only to determine whether you own the GitHub organization an installation belongs to.
  • For enterprise single sign-on, your organization identifier and the claims your identity provider sends, which may include your name, email, and verification status.

Account and workspace information

We store your email address, name, optional profile image URL, account creation and last-login times, workspace memberships and roles, team memberships, budgets, and workspace settings. If you are invited to a workspace, we store the email address the invite was sent to and the invite status. Workspace owners and administrators can see the names, email addresses, roles, spend, token usage, budgets, teams, and last-active times of people in their workspace.

Billing information

Payments are processed by Stripe. We do not receive or store your full card number or CVC. We store a payment method reference, card brand, last four digits, expiry date, billing transactions, credit balance and ledger entries, invoices, and your Stripe customer identifier. Stripe receives your email address and payment details when you make a purchase and handles tax calculation where applicable.

Usage and telemetry

When you use the command-line interface, we collect:

  • Session identifiers, machine identifier, model, token counts, cost, duration, and timestamps.
  • A short prompt excerpt (up to 160 characters), an opening prompt (up to 2,000 characters), message counts, tools used, files read (up to eight paths), files written, and a generated session name and summary.
  • Run telemetry such as CLI version, agent type, a hashed repository identifier, primary language, token and cache counts, routing parameters, model tiers, pass or fail status, failure signature, and latency.

We do not store the full session transcript. The transcript stays on your machine. Live session data relayed through the dashboard is held in process memory only and is not written to our database. Relay control commands are logged without prompt text.

Cloud sandboxes

When you start a cloud sandbox, we store the repository, base branch, branch, session identifier, status, compute time, charges, pull request number and URL, and a bounded failure reason. The task prompt you enter is stored only until the run starts and is then cleared. The sandbox workspace is destroyed when the run ends, and the sandbox record does not contain your repository contents, the diff, or the transcript.

GitHub pull request reviews

When you connect a repository, we store the installation identifier, account login, repository selection, repository name, and whether the repository is private. When a review runs, we store the repository, pull request number, title, author, commit, base branch, status, findings, and cost. We use a short-lived GitHub token scoped to the repositories the GitHub App can access. Review checkouts are deleted when the review finishes.

Communications and feature requests

If you contact us, submit a feature request, or join a waitlist, we collect the information you provide, including your name, email address, and message. Feature requests may be visible to your workspace and to our team, along with the display name you used.

Technical information

Our hosting and infrastructure providers may log IP addresses, request metadata, and error information as part of operating and securing the Service. Our application does not read or store your IP address or user agent. We do not use third-party analytics, advertising, or session-replay tools in the Service.

2. Cookies and local storage

We use these cookies and browser storage:

NamePurposeLifetimeNotes
Auth.js sessionKeeps you signed in and stores your session claims, including a GitHub OAuth token for GitHub sign-in.Session, managed by Auth.jsEncrypted, HttpOnly
midas-wsRemembers your last workspace.400 daysNot HttpOnly
midas-promoRemembers a promotional code so it can be attached at checkout.30 daysNot HttpOnly, scoped to .trymidas.io
midas-setup-doneRemembers onboarding progress.400 daysNot HttpOnly, includes a local account identifier
midas_github_oauth_stateProtects the GitHub authorization flow against CSRF.10 minutesHttpOnly
midas_github_workspaceRemembers the workspace during GitHub setup.10 minutesHttpOnly
midas-theme (local storage)Remembers your light or dark theme.Until clearedStored in your browser
midas-public-profile (local storage)Remembers a local profile preference.Until clearedPublic profiles are not currently available
midas-feature-request-draft (session storage)Preserves an unsent feature request.Until the tab closesStored in your browser

The session, security, and workspace cookies are necessary to provide the Service. The promotional cookie stores only a campaign code and is not used to track you across sites. We do not use cookies for advertising, analytics, or cross-site tracking. We do not currently display a cookie consent banner.

3. How we use information

We use personal information to:

  • Create and authenticate accounts, and manage workspaces, memberships, and roles.
  • Provide the Service, including running the agent, executing sandboxes, reviewing pull requests, and displaying session history.
  • Meter usage, calculate charges, process payments, manage credits and auto-reload, and provide invoices.
  • Send service messages, including invites, balance and usage notifications, and security notices.
  • Provide support and respond to requests.
  • Secure the Service, detect abuse, enforce limits, and investigate incidents.
  • Comply with legal, tax, accounting, and regulatory obligations.
  • Understand aggregate usage and improve the Service. We do not use your code or prompts to train models.

We rely on the following legal bases where the General Data Protection Regulation applies:

  • Contract: to provide the Service and process payments.
  • Legitimate interests: to secure the Service, prevent abuse, improve the product, and communicate about the Service, balanced against your rights.
  • Legal obligation: to keep records for tax, accounting, and legal purposes.
  • Consent: where we ask for it, such as for optional marketing communications. You can withdraw consent at any time.

We do not sell personal information, and we do not share it for cross-context behavioral advertising.

4. AI model providers and zero data retention

The Service sends your prompts, relevant code and repository context, and tool output to third-party AI model providers so they can generate a response. This is necessary to provide the Service.

We route requests through OpenRouter to upstream model providers. We configure zero-data-retention routing where the selected model supports it, so the provider does not retain the request. Not every model has a zero-retention endpoint. Where a model does not, the upstream provider's own retention terms apply. We do not use your code or prompts to train models.

We store only the limited session metadata described in Section 1, including a short opening prompt and prompt excerpt used to show session history. We do not store the full transcript. Sandbox task prompts are cleared when the run starts, and review checkouts are deleted after the review.

Do not submit information to the Service that you are not permitted to send to a third-party AI provider.

5. How we share information

We share personal information with service providers that help us operate the Service:

  • Stripe: payment processing, invoices, tax calculation, and fraud prevention.
  • OpenRouter and upstream AI model providers: inference for agent requests, sandboxes, reviews, and session summaries.
  • GitHub: repository access, pull request reviews, and GitHub App installation.
  • Google: sign-in when you choose Google.
  • Your identity provider: single sign-on and SCIM provisioning when your organization configures it.
  • Resend: transactional email, including workspace invites and notifications.
  • Google Cloud: hosting, database, storage, networking, and logging infrastructure.
  • Redis: internal job queues. Queue messages contain job identifiers, not customer payloads.

We may also disclose information:

  • To your workspace owners and administrators, as described in Section 1.
  • To our professional advisers, auditors, and insurers.
  • To a successor in connection with a merger, acquisition, financing, reorganization, or sale of assets.
  • When required by law, legal process, or a regulator, or when we believe disclosure is necessary to protect rights, safety, or security.
  • With your consent or at your direction.

We require service providers to protect personal information and to use it only for the purposes we specify, consistent with their own terms and applicable law.

6. International transfers

We operate the Service in the United States, primarily in Google Cloud's us-central1 region. If you access the Service from another country, your information will be transferred to and processed in the United States, where privacy laws may differ from those in your country.

Where required, we rely on appropriate transfer mechanisms, including Standard Contractual Clauses and the UK International Data Transfer Addendum, and we assess the protections available in the destination country.

7. Retention

We keep personal information for as long as needed to provide the Service and for the purposes described in this Policy. Specific retention practices:

  • Account, workspace, and membership data: while your account is active and for a reasonable period afterward.
  • Session metadata, prompt excerpts, summaries, run telemetry, sandbox records, and review records: retained as part of your account history. Account deletion does not erase this history.
  • Billing ledgers, invoices, and tax records: retained as required for accounting, tax, and fraud prevention. These records are append-only.
  • Audit logs: retained for security and compliance.
  • Email delivery records: deleted 30 days after delivery or final failure.
  • Invite rate-limit records: deleted after 2 days.
  • Device authorization codes: expire after 15 minutes.
  • Single sign-on handoff codes: expire after 120 seconds.
  • Live session relay data: held in memory only and not persisted.
  • Sandbox task prompts: cleared when the run starts.
  • Review checkouts: deleted when the review finishes.
  • Waitlist and contact submissions: retained until you ask us to delete them, unless we need them for a legal purpose.

When you delete your account, we anonymize your user record, delete your linked identities and memberships, revoke your CLI tokens, clear your saved card and auto-reload settings, and soft-delete your personal workspace. We retain billing, usage, session, sandbox, review, and audit records as described above. We may also retain information in backups for a limited period and in records we are legally required to keep.

8. Security

We use administrative, technical, and physical safeguards designed to protect personal information, including:

  • Encryption in transit using TLS.
  • Encryption at rest for our database and storage.
  • Application-level encryption for stored secrets, including GitHub tokens, identity provider client secrets, and invite tokens.
  • Hashing of CLI tokens, invite tokens, device codes, single sign-on codes, and SCIM tokens.
  • Private networking and least-privilege access controls.
  • Keeping full payment card data out of our systems by using Stripe's hosted payment fields.

No method of transmission or storage is completely secure. We cannot guarantee absolute security. If we become aware of a breach affecting your personal information, we will notify you and any regulators as required by law.

9. Your rights

Depending on where you live, you may have the right to:

  • Access the personal information we hold about you.
  • Correct inaccurate information.
  • Delete your information, subject to legal retention obligations.
  • Receive a copy of your information in a portable format.
  • Restrict or object to certain processing.
  • Withdraw consent where processing is based on consent.
  • Lodge a complaint with a supervisory authority.

You can exercise many of these rights in the dashboard, including updating your profile, managing workspace memberships, and deleting your account. For other requests, contact us at hello@trymidas.io with "Privacy request" in the subject line. We may need to verify your identity before we act. We will respond within the time required by applicable law.

We do not currently offer a self-serve data export tool. You can request an export, and we will provide the information we can reasonably compile.

If you are in California, you also have the right to know what categories of personal information we collect, disclose, and retain, the right to delete, the right to correct, the right to opt out of sale or sharing (we do not sell or share personal information for cross-context behavioral advertising), and the right not to be discriminated against for exercising your rights. You may designate an authorized agent to make a request on your behalf.

If your organization is the controller of your information, please contact your organization first. We will assist it in responding to your request.

Enterprise customers can request a data processing addendum by contacting hello@trymidas.io.

10. Children

The Service is intended for professional use by people who are at least 18 years old. We do not knowingly collect personal information from children. If we learn that we have collected information from a child, we will delete it.

11. Changes to this Policy

We may update this Policy. If we make a material change, we will notify you by email or in the dashboard before the change takes effect, unless the change is required sooner by law. The updated Policy will be posted at trymidas.io/privacy with a new "Last updated" date.

12. Contact

For privacy questions or requests:
[LEGAL ENTITY NAME]
[ADDRESS]
hello@trymidas.io