Checkpoint handles API keys, OAuth tokens and the contents of your tickets, so we take security reports seriously.

Please don't open a public issue. Report privately through GitHub:

➡️ Report a vulnerability (Security → Advisories → Report a vulnerability)

Include what you found, how to reproduce it, the version affected, and the impact you think it has. You'll get an acknowledgement within 3 business days and a status update within 7 days. We'll agree a disclosure date with you and credit you in the release notes (unless you'd rather not be named).

Security fixes go into the latest release. Please update before reporting.

In scope — for example:

• Leaking API keys or OAuth tokens (Keychain handling, logs, crash output, network)
• Checkpoint writing to Jira or Linear, or exposing write tools to a model, without explicit user action
• Codebase access escaping the folder the user chose (path traversal, symlinks)
• The OAuth loopback listener being reachable from other machines, or accepting forged callbacks
• Prompt injection in ticket or code content that causes actions beyond producing a plan
• Sandbox or entitlement weaknesses in release builds

Out of scope:

• Vulnerabilities in Atlassian, Linear, AI providers or their MCP servers (report to them)
• A model producing an inaccurate test plan
• Issues requiring a compromised Mac or a malicious local user with your account

• Keys and tokens live in the macOS Keychain, never in plain files.
• Read-only trackers: only read tools are exposed for Jira; Linear uses its read-only MCP endpoint and read scope.
• Codebase access (scenarios) is read-only and confined to a user-chosen folder via a security-scoped bookmark.
• Sandboxed app with network client, a loopback-only listener for sign-in, and read-only access to user-selected folders.
• No Checkpoint servers: the app talks directly to your tracker and your chosen AI provider.