- Implemented a new workflow step to clone the SDK repository and update the OpenAPI specification.
- Configured Git user details for the sync operation and added a commit message format that includes source and update timestamp.
- Ensured successful synchronization of OpenAPI documentation to the SDK repository.
- Changed the path of the `migrate-auth-secret` script from the root directory to the `scripts` folder for better organization and clarity in the project structure.
- Added a regex validation for branch names in Bitbucket, Git, Gitea, GitHub, and GitLab provider schemas to ensure consistent and valid branch formats.
- Refactored the branch validation logic to improve readability and maintainability across the schemas.
- Enhanced input integrity by ensuring all provider schemas adhere to the same branch name validation rules.
- Added a new script `migrate-auth-secret.ts` to facilitate the migration of 2FA secrets when changing the BETTER_AUTH_SECRET.
- Updated `package.json` to include a command for running the migration script.
- Refactored the handling of BETTER_AUTH_SECRET to improve security by removing the hardcoded default and introducing a fallback mechanism using environment variables or Docker secrets.
- Updated the authentication logic to utilize the new `betterAuthSecret` function for retrieving the secret.
- Introduced a new validation schema for registry URLs, ensuring they conform to a specific format (hostname[:port]) and disallow shell metacharacters.
- Updated the `createSchema`, `apiCreateRegistry`, and `apiTestRegistry` functions to utilize the new registry URL validation.
- Improved security and input integrity for registry URL fields.
- Updated the `removeRegistry` function to escape the registry URL during logout to prevent command injection vulnerabilities.
- Updated the `destinationPath` field in the upload file schema to include a regex validation, ensuring only alphanumeric characters, dots, dashes, underscores, and forward slashes are allowed.
- Added a corresponding regex check in the `uploadFileToContainer` function to validate the destination path before processing, improving input integrity and preventing errors.
- Updated the WebSocket server to encode the log path in base64 before executing the tail command on the remote server.
- Added validation to ensure the directory name adheres to a specified regex pattern, improving input integrity for directory paths.
- Modified the `writeConfigRemote` function to encode the Traefik configuration in base64 before saving it to the remote YAML file.
- This change ensures that the configuration is correctly handled and prevents potential issues with special characters in the config.
- Introduced a regex validation for branch names in Bitbucket, Git, Gitea, GitHub, and GitLab provider schemas to ensure valid branch formats.
- Updated the corresponding schemas to include the new validation rule, enhancing input integrity and preventing potential errors.
- Added a utility for branch validation in the server utils.
- Updated the schedule form schema to include an optional 'description' field.
- Enhanced the form to allow users to input a description for each schedule.
- Modified the schedule display component to show the description if available.
- Added a database migration to include the 'description' column in the schedule table.
Introduced an optional 'isolated' boolean property in the CompleteTemplate interface to manage isolated deployment settings. Added tests to verify default behavior (isolated=true) and explicit settings (isolated=true/false) in the deployment configuration.
This change enhances template flexibility for deployment configurations.
Templates using network_mode: host (e.g. cloudflared) can now declare
isolated = false in their [config] section to prevent Dokploy from
injecting networks into the compose, which would cause a Docker error.
Default behavior (isolated = true) is unchanged for all existing templates.
Fixes#4366
Add 10s AbortSignal timeout to all template fetch calls so they fail
cleanly instead of hanging indefinitely when templates.dokploy.com is
unreachable. Add try/catch to getTags endpoint which was missing error
handling, causing a 500 instead of returning an empty list.
Closes#4282
Replace permanent (301) redirects with temporary (302) redirects across
all pages that check authentication state in getServerSideProps.
Permanent redirects are cached by the browser indefinitely, causing a
bug where users had to manually refresh after login: the browser would
cache the unauthenticated redirect (dashboard → login page) and replay
it even after a successful login, preventing navigation to the dashboard.
Fixes#4220
Closes#3909
The previous 10s interval triggered a DB query (SELECT 1) 360 times/hour,
keeping Node.js heap under constant pressure and preventing the GC from
reclaiming memory efficiently. Increasing to 30s reduces query load by 3x.
Also adds start-period=60s to avoid false failures during startup.
- Added a maximum length constraint of 255 characters for the email input field.
- Updated validation schema to include a message for exceeding the maximum length.
- Added `sendInvitationEmail` function to send invitation emails when a new organization is created in the cloud environment.
- Updated email template to enhance the invitation message and included a direct link for users to accept the invitation.
- Refactored email sending logic in the user router to utilize the new invitation email rendering function.
- Improved organization invitation email design for better user experience.
- Implemented checks in the WebSocket server setups for Docker container logs, terminal, and deployment logs to ensure users can only access resources associated with their active organization.
- Enhanced security by closing WebSocket connections if the organization ID does not match the session's active organization ID.
Prevents owner/admin users of one organization from accessing servers,
destinations, and Docker Swarm join tokens belonging to other organizations
by validating organizationId on all endpoints that accept serverId or
destinationId as direct input.
- cluster: validate serverId org on getNodes, addWorker, addManager, removeWorker
- deployment: validate serverId org on allByServer
- backup: validate destinationId + serverId org on listBackupFiles
- volume-backups: validate destinationId + serverId org on restoreVolumeBackupWithLogs
- wss: validate server org on docker-container-logs, docker-container-terminal,
listen-deployment, and terminal WebSocket handlers
- auth: fix TypeScript type for API key metadata parsing
- Added comprehensive permission checks for creating, updating, and deleting schedules based on user roles (owner/admin) and schedule types (server/dokploy-server).
- Implemented restrictions for cloud users to prevent managing host-level schedules and changing schedule types.
- Improved access control for server-level schedules to ensure users can only manage schedules associated with their organization.
- Adjusted the Card component to have a minimum height of 85vh for better visual consistency.
- Ensured the inner div has a full height to enhance the layout structure.
- Replace individual project and server queries with a consolidated homeStats query to streamline data retrieval for the dashboard.
- Update the ShowHome component to utilize homeStats for displaying project, environment, application, and service counts, along with their status breakdown.
- Enhance data handling for user permissions to ensure accurate statistics based on user access levels.
- Home: 4 KPI cards (projects, services, deploys/7d, status list),
server column with icon in recent deployments, empty state with
icon, dashboard card frame to match other pages.
- Include libsql in project services count sort.
- Fix bulk actions in environment page: libsql was missing from
start, stop, move, delete and deploy handlers.
Adds a new /dashboard/home landing with welcome header, KPI cards
(deploys/24h, build, CPU, memory) and a recent deployments list.
Home is now the post-login landing and the destination for permission
fallback redirects across the app. Projects remains accessible from
the sidebar.
Update the version retrieval command in the GitHub Actions workflow to strip the 'v' prefix from the version number in package.json. This change ensures that the version format is consistent for downstream processes.
Modify the GitHub Actions workflow to clone the MCP and CLI repositories into temporary directories instead of the current directory. This change improves the organization of the workflow and ensures that the latest OpenAPI specification is correctly referenced during the synchronization process.
Enhance the GitHub Actions workflow by adding a workflow_dispatch trigger, allowing manual execution of the version synchronization process. This provides greater flexibility in managing version updates for MCP and CLI repositories.
Update the GitHub Actions workflow to bump the version in package.json after installing dependencies, ensuring that the version is not overwritten by pnpm install. This change enhances the reliability of version synchronization for both MCP and CLI repositories.
Change the logger's disabled property to be dependent on the NODE_ENV variable, ensuring logging is disabled in production for improved performance and security.
- Add `sendOnSignIn: true` to emailVerification config so unverified users
receive a new verification email when they attempt to sign in
- Create styled verification email template matching the invoice email design
- Extract `sendVerificationEmail` helper to keep auth.ts clean
- Show friendly message on login when email is not verified
Enhance the TagSelector component to display a message when no tags are created, prompting users to add tags. This improves user experience by providing clear feedback in the UI.
Update the GitHub Actions workflow to include regeneration of tools from the latest OpenAPI specification and ensure the latest openapi.json is copied to the CLI repository. This improves the consistency and accuracy of the versioning and API documentation across both repositories.
Implement a GitHub Actions workflow to automatically sync the version from the Dokploy repository to the MCP and CLI repositories upon release. This includes cloning the repositories, updating the package.json version, and committing the changes with relevant metadata, ensuring consistent versioning across platforms.
Implement workflows to sync the OpenAPI specification to both the MCP and CLI repositories. This includes cloning the repositories, updating the openapi.json file, and committing the changes with relevant metadata. The process ensures that the OpenAPI documentation is consistently updated across multiple platforms.
Integrate a new component, ShowContainerNetworks, to display network details for each container in the dashboard. This includes a dialog that shows network information such as IP address, gateway, and MAC address, enhancing the container management capabilities.
Add a new "View Mounts" action to the container dropdown that displays
volume and bind mounts in a formatted table (type, source, destination,
mode, read/write). Also add "View Config" and "Terminal" actions to the
compose containers tab which previously only had logs and lifecycle actions.
Right-click on service cards to quickly Start, Deploy, Stop, or Delete
a service without navigating into it. Uses shadcn/ui ContextMenu
component built on @radix-ui/react-context-menu. Delete action shows
a confirmation dialog. LibSQL services are excluded since they lack
standard mutation endpoints.
Change the audit action from "kill" to "stop" for the containerKill function to better reflect the operation being performed. This aligns the logging with the intended action and improves clarity in audit records.
Change restartContainer, startContainer, stopContainer, and killContainer
endpoints to use service.read instead of docker.read so members with
access to the compose can use container lifecycle actions.
Add a Containers tab to the compose service page that lists all
containers with their state, status, and container ID. Each container
has a dropdown menu with lifecycle actions: View Logs, Restart, Start,
Stop, and Kill.
- Add containerStart, containerStop, containerKill functions to docker service
- Add corresponding tRPC procedures with server ownership checks and audit logging
- Update containerRestart to support remote servers via serverId
- Create ShowComposeContainers component with table view and action menu
- Add Containers tab between Deployments and Backups, gated by docker.read permission
- Simplified code formatting for better readability in various components.
- Updated job scheduling functions to handle errors gracefully, ensuring that failures in scheduling do not disrupt the overall process.
- Enhanced logging for better traceability of job scheduling issues.
These changes improve code maintainability and user experience by providing clearer error messages and more organized code structure.
- Added customer update fields for automatic name and address handling during subscription creation.
- Enabled billing address collection and tax ID collection for improved compliance and billing accuracy.
These changes enhance the Stripe payment process by ensuring necessary customer information is captured and managed effectively.
- Introduced a new feature allowing users to enable or disable invoice email notifications in the billing settings.
- Implemented email notifications for successful invoice payments and payment failures, enhancing user communication regarding billing.
- Updated the database schema to include a new column for storing user preferences on invoice notifications.
- Added corresponding email templates for invoice notifications and payment failure alerts.
These changes improve user experience by keeping users informed about their billing status and actions required.
When both stripPath and internalPath are configured, addPrefix was pushed
before stripPrefix causing incorrect path rewriting (e.g. /app/v2/public/api
instead of /app/v2/api). Traefik executes middlewares in array order, so
stripPrefix must come first.
Closes#4061
When users set a custom docker compose command without the -p flag,
Docker Compose defaults to using the directory name (code) as the
project name. If the custom command is later removed, Dokploy uses
-p appName, creating a new stack while the old one remains running.
Injecting COMPOSE_PROJECT_NAME=appName into the .env ensures the
project name is always consistent regardless of the command used.
Closes#4019
Commands were chained with && so if the project directory was missing,
cd would fail and docker compose down would never execute — leaving
containers and volumes running. Use semicolons to run each command
independently, matching the existing stack deletion pattern.
Closes#4064
- Updated the AnalyzeLogs component to display a message and button for configuring AI providers when none are available, improving user guidance.
- Added a link to the settings page for easy access to AI provider configuration.
- Integrated new icon for the configuration button to enhance UI clarity.
These changes improve the user experience by ensuring users are informed about the need to set up AI providers for log analysis.
- Updated the `getContainerLogs` function to accept either an application name or container ID, improving flexibility in log retrieval.
- Simplified the command execution logic by consolidating the remote and local execution paths.
- Added a new parameter to directly use container IDs, streamlining the process for users.
These changes enhance the usability of the logging feature, allowing for more efficient access to container logs.
- Implemented a new `readLogs` procedure across various routers (application, compose, libsql, mariadb, mongo, mysql, postgres, redis) to enable users to retrieve logs from containers.
- Each procedure includes input validation for parameters such as `tail`, `since`, and `search`, ensuring robust access control and authorization checks.
- Enhanced the `getContainerLogs` service to support fetching logs from both Docker containers and services, improving the logging capabilities of the application.
This feature enhances observability and troubleshooting for users by providing direct access to container logs.
- Corrected the API URL for Z.AI by removing the trailing slash.
- Modified the AI router mutation to include context and added access control to ensure users can only access their organization's AI settings.
These changes improve the accuracy of the API integration and enhance security by enforcing organizational access restrictions.
- Introduced the AnalyzeLogs component for analyzing logs using AI, allowing users to select AI providers and view analysis results.
- Integrated AnalyzeLogs into the ShowDeployment and DockerLogsId components, enabling log analysis for both build and runtime contexts.
- Updated the AI router to include a new endpoint for log analysis, which processes logs and returns structured insights.
- Enhanced the AI provider selection logic to support new providers, including Z.AI and MiniMax.
This feature enhances the user experience by providing actionable insights from logs, improving troubleshooting and operational efficiency.
Add allForApps endpoint that returns only sshKeyId and name using protectedProcedure instead of withPermission, so members can select SSH keys in the git provider dropdown without needing access to the SSH Keys management panel.
closes#4069
Move server.listen() before the initialization block so the HTTP server
is already responding when Docker healthchecks begin. Previously, slow
operations like SMTP timeouts in sendDokployRestartNotifications() could
block the server from listening, causing healthcheck failures and
container restarts.
Closes#4049
The catch block was swallowing the real error from the ntfy server,
making it impossible to diagnose connection failures (e.g. SSL, DNS,
auth issues). Now the underlying error message is included in the
tRPC error response.
Closes#4047
- Change compose.loadServices permission from service:create to service:read
since loading services from a compose file is a read-only operation
- Add saveEnvironment endpoint to compose router with envVars:write permission
- Update show-environment.tsx to use saveEnvironment mutations instead of
generic update mutations for all service types (compose, databases)
Closes#4052
The actions array in Slack attachments requires Interactive Components
to be configured on the Slack app, which causes notifications to fail.
Replaces with a Details field using mrkdwn hyperlink syntax and adds
mrkdwn_in to ensure the link renders as clickable.
Closes#4053
writeDomainsToCompose reads the compose file in Node.js before the
shell script runs, so patches applied as shell commands were being
overwritten by the stale pre-patch content.
Split patch execution into a separate step that runs before
getBuildComposeCommand, so the file is already patched when Node.js
reads it for domain injection.
Also added missing patch support to rebuildCompose which was skipping
patches entirely on redeploys.
Closes#4113
The naive split(":").pop() approach treated the port number in
registry URLs (e.g. registry:5000/image) as the image tag.
Now uses lastIndexOf(":") and checks if the suffix matches a port
followed by a path (digits + slash), consistent with extractImageName.
Closes#4082
- Normalize emails with toLowerCase().trim() before comparing
- Wrap getUserByToken in try/catch since it throws TRPCError on miss,
rethrow as APIError for consistent error responses
Also checks that the invitation is not expired and has not already been
used before allowing account creation.
Co-Authored-By: Claude Opus 4.6 (1M context) <[email protected]>
- Updated the authentication process to check if the email of the user matches the email associated with the invitation token.
- Improved error handling for cases where the user is not found or the email does not match the invitation.
- Added `isEnterpriseCloud` field to the user schema to enhance user differentiation for enterprise cloud services.
- This change supports the ongoing updates to billing and subscription management for enterprise users.
- Introduced `isEnterpriseCloud` boolean field in the user schema to differentiate enterprise users.
- Updated billing UI to display specific information for enterprise cloud users, including a dedicated section for managing subscriptions.
- Modified API webhook logic to handle subscription updates and server management based on the `isEnterpriseCloud` status.
- Introduced `accessedServers` field in user permissions schema and member table.
- Implemented server access validation across various API routers to ensure users can only access permitted servers.
- Added a new query to fetch accessible server IDs based on user roles and licenses.
- Updated UI components to support server selection in user permissions.
Resolve conflicts:
- Integrate credentials-based user provisioning with canary changes
- Use withPermission("member", "create") instead of adminProcedure
- Adopt standardSchemaResolver, inviteMember mutation, and custom roles from canary
- Restrict credentials flow to non-cloud environments
- Enhanced the `formatSizeValue`, `formatMemUsage`, and `formatIOValue` functions to handle edge cases more effectively.
- Updated regex and condition checks to ensure proper parsing of input strings, improving overall reliability in formatting size values.
- Introduced a tabbed layout in the ShowSwarmOverviewModal to separate the overview and containers views.
- Added ShowSwarmContainers component to the containers tab, improving the organization of information.
- Integrated Card component for better styling and presentation of the containers section.
- Introduced new components for handling various empty states in the Swarm containers dashboard, including `SwarmNotAvailable`, `ServicesError`, `NoServices`, and `NoRunningContainers`.
- Added `SummaryCards` component to display key metrics such as node count, down nodes, service count, and running containers.
- Enhanced the `ShowSwarmContainers` component to integrate the new empty states and summary cards, improving user feedback and overall experience.
- Introduced a new optional `icon` field to the application schema, allowing for icon uploads.
- Implemented validation to ensure the icon size does not exceed 2MB, enhancing data integrity.
- Updated the ShowIconSettings component to include a dialog for icon selection and upload.
- Added functionality to handle file uploads, including validation for file types and sizes.
- Implemented icon removal feature within the dialog.
- Refactored icon selection logic to improve user experience and maintainability.
- Adjusted the application page to integrate the updated ShowIconSettings component.
- Added `dompurify` for sanitizing SVG icons to prevent XSS vulnerabilities.
- Introduced `simple-icons` for a collection of SVG icons, enhancing the icon selection feature.
- Updated the `ShowIconSettings` component to utilize the new icon management logic.
- Removed the obsolete `icons.json` file and replaced it with a new `bundled-icons.ts` file for better structure and maintainability.
- Adjusted related API and component files to accommodate the new icon handling approach.
Mauricio Siu
