Your AI team, working in Railway
Railway is a deployment platform for building and shipping applications with instant deployments, automatic scaling, and managed infrastructure.
Start your companyWhat your team can do in Railway
372 tools to work with, connected with an API key.
Add a workspace member to an access group.
Add a feature flag for a user
Add a feature flag for a project
Add a workspace member to a project with a specific role. The user must already be a member of the project's workspace.
Add a feature flag for a service
Add a deploy source to a workspace policy allowlist.
Approves a deployment.
Archive a cloud agent: hide it from lists that ask for unarchived agents. Its machine, sessions and history are untouched. Owner only.
Attach a project to an access group.
Build a sandbox template into a bootable checkpoint.
Stop a task. Before its prompt is sent it never starts; after, only its own run is aborted. The task ends failed with "The task was cancelled". A finished task is left as is.
Cancel a Railway deployment that is queued or building, preventing that in-progress deployment from completing. This does not stop a running deployment and requires confirm=true.
Cancels an in-progress PITR enable/disable rollout on an HA database cluster and resets its progress so the UI re-reads from the live cluster. No-ops when nothing is running.
Cancel scheduled deletion of a project
Check whether any Railway service in an environment uses legacy static egress rather than high-availability static egress. Returns a Boolean suitable for identifying environments needing HA migration. Read-only: does not enable, disable or migrate egress.
Check whether a proposed private-network endpoint name prefix is available in a Railway environment and private network. Returns an availability boolean; false means unavailable, not an execution error. Does not reserve the name, create an endpoint or change DNS.
Check whether a Railway service domain is available. Returns an availability Boolean and Railway's explanatory message. A false result is a normal lookup result, not a transport error. This read-only check does not reserve, create or change a domain and does not guarantee future availability.
Check whether the upstream repository for a Railway service instance has an update available. Returns Railway's Boolean availability result; false is a valid result. This is an upstream repository update check, not an authorization or configuration-editability check. Does not pull changes, redeploy or modify the service.
Claims a project.
Claim a build requested from @RailwayAgent on X into workspaceId. The caller's connected X account must be the one that requested it.
Forget the box `ssh dev.new` lands you in; the next connection creates a new one.
Clear all egress gateway associations for a service instance
Clears the recorded progress of a finished PITR enable/disable rollout so the Backups page starts fresh. A rollout still in progress cannot be cleared.
Undo an active auto-update skip for a service instance, applied immediately (no config staging, no redeploy). The recurring maintenance window and update policy are untouched β the next scheduled occurrence fires normally again.
Duplicates an existing template
Collect a bounded, best-effort Railway deployment evidence bundle from live-validated deployment, snapshot, event, service-instance, log, and metric sources. Results may be partial and report completeness and errors for each source. This tool performs no AI reasoning or root-cause diagnosis and does not modify code or open pull requests.
Commit existing staged changes for a Railway environment, completing a staged STAGE_TEMPLATE or create-environment workflow. This irreversible update can start deployments unless skip_deploys=true and requires confirm=true.
Commit an explicit Railway patch for bucket creation, updates, or deletion and the environment private-network enabled state. Existing staged changes are not committed or replaced, and this tool does not change services, variables, or volumes. Bucket deletion is irreversible and requires confirm_destructive=true.
Confirm the transfer of project ownership
Connect a service to a source
Count all published Railway templates in the public catalog. Returns one total without listing templates. Accepts no filters, workspace/project IDs or pagination.
Create an access group in a workspace.
Create a project-level Railway object-storage bucket and return its identity, ownership, and timestamps. This single mutation does not apply an environment configuration patch or select a storage region; environment deployment may require separate configuration. Returns metadata only, without S3 credentials.
Create a persistent Railway cloud agent in an environment, optionally from a checkpoint or repository source, with optional region and public code endpoint. Returns identity and lifecycle metadata only. This action does not accept raw variable values. Creation can provision a running machine.
Checkpoint the agent's disk. Poll cloudAgentCheckpoint until SUCCEEDED.
Send lightweight agent feedback to Railway. Returns the report ID after delivery.
Invoke a deployment instance execution.
Creates a deployment trigger.
Create either a generated railway.app domain or a caller-owned custom domain. Custom-domain responses can include DNS setup records and Railway's verification token and state when available.
Create a new egress gateway association for a service instance
Create an empty Railway environment or clone an existing environment, with explicit initial-deployment and change-staging behavior.
Create an integration for a project
Creates a new job application.
Create a new notification rule
Create an observability dashboard
Deprecated by Railway: Removed. Private networks and endpoints are managed by the platform; enable private networking on the environment instead. Removed. Private networks and endpoints are managed by the platform; enable private networking on the environment instead.
Deprecated by Railway: Removed. Private networks and endpoints are managed by the platform; enable private networking on the environment instead. Removed. Private networks and endpoints are managed by the platform; enable private networking on the environment instead.
Creates a new plugin in Railway for the specified project. Use this action when you need to add a custom plugin to a Railway project, such as for integrating external services or extending project functionality. The action requires a valid project ID (found in the Railway dashboard URL) and a descriptive name for the plugin. The plugin will be created with default settings and can be further configured after creation.
Create a Railway project in an accessible workspace and return its project and default-environment IDs. Optionally connect a repository branch for deployment and configure the runtime, public visibility, and pull-request deployment settings.
Create an invitation for a project
Create a DNS record for a Railway domain
Forward an address on this domain to an external inbox
Create a Railway sandbox in an environment, optionally forking a running sandbox or building/restoring a template. Configure resources, idle timeout, networking and public domains. Returns lifecycle and domain metadata; does not wait for RUNNING. Environment-variable value inputs are not supported.
Capture a running sandbox's current disk into a reusable, named checkpoint. Synchronous: the checkpoint is ready when this returns.
Create an empty, repository-backed, image-backed, or template-backed service in a Railway project.
Register a new signal for an owner scope.
Deprecated by Railway: Use staged changes and apply them. Creating a TCP proxy with this endpoint requires you to redeploy the service for it to be active. Creates a new TCP proxy for a service instance.
Create a new trusted domain for this workspace
Create a persistent Railway volume and optionally attach it to one service and environment. This action creates billable persistent state and does not clean it up; deletion is a separate DELETE_VOLUME action that requires workspace-admin authorization when using OAuth.
Start an asynchronous backup of a Railway volume instance, optionally assigning a name (default Manual). Supply the volume instance ID, not its parent volume ID. Manual backups require a Pro workspace. Returns a nullable workflow ID; acceptance does not confirm backup completion.
Delete an access group. This is destructive and requires confirm=true.
Delete a Railway cloud agent by ID. This is a destructive operation; use only for an agent you intend to remove and requires confirm=true. Returns the provider's Boolean result: true confirms success, while false does not confirm deletion.
Delete a checkpoint. This is destructive and requires confirm=true.
Permanently delete a completed Railway deployment from deployment history and stop any remaining project usage from it. This does not delete the service, is irreversible, and requires confirm=true.
Deletes a deployment trigger. This is destructive and requires confirm=true.
Permanently delete one generated Railway domain or custom domain by ID after explicit confirmation.
Permanently delete one Railway environment and its deployments. This action is irreversible, requires confirm=true, and requires workspace-admin authorization when using OAuth. Reconnect with a separately selected workspace:admin grant if Railway returns Not Authorized.
Delete an integration for a project This is destructive and requires confirm=true.
Delete a notification rule This is destructive and requires confirm=true.
Deprecated by Railway: Plugins are deprecated Deletes a plugin. This is destructive and requires confirm=true.
Deprecated by Railway: Removed. Private networks and endpoints are managed by the platform; enable private networking on the environment instead. Removed. Private networks and endpoints are managed by the platform; enable private networking on the environment instead. This is destructive and requires confirm=true.
Deprecated by Railway: Removed. Private networks and endpoints are managed by the platform; enable private networking on the environment instead. Removed. Private networks and endpoints are managed by the platform; enable private networking on the environment instead. This is destructive and requires confirm=true.
Permanently delete one Railway project and its resources. This action is irreversible and requires explicit confirm=true. OAuth requires a workspace:admin grant for the selected workspace and a user with the Admin role.
Delete an invitation for a project This is destructive and requires confirm=true.
Deletes a project with a 48 hour grace period. This is destructive and requires confirm=true.
Delete a DNS record for a Railway domain This is destructive and requires confirm=true.
Stop forwarding an address on this domain This is destructive and requires confirm=true.
Delete a sandbox checkpoint. This is destructive and requires confirm=true.
Permanently delete a Railway service; requires confirm=true. With environment_id, delete only in that environment; the service itself is deleted once no instances remain. Without it, delete in every environment that is not a fork. Requires member access to the selected project or workspace.
Delete a signal and its audit log. This is destructive and requires confirm=true.
Deletes a TCP proxy by id This is destructive and requires confirm=true.
Delete a Railway template by ID, optionally in a workspace context. This is a destructive operation requiring confirm=true and permission to manage the template. Returns the provider's Boolean result: true confirms success; false does not confirm deletion.
Delete a trusted domain This is destructive and requires confirm=true.
Delete a Railway environment variable. This action uses the `variableDelete` GraphQL mutation to permanently remove an environment variable from a project, environment, or service. This action is irreversible β once deleted, the variable cannot be recovered.
Permanently delete a persistent volume and all its associated data from Railway. Use when you need to remove a volume that is no longer needed. This action is irreversible β the volume and all its data will be permanently deleted and cannot be recovered once removed. Use this action when: - Cleaning up unused volumes to reduce storage costs - Removing volumes associated with deprecated environments - Deleting volumes before removing a project
Deletes volume instance backup This is destructive and requires confirm=true.
Delete a Railway workspace and all data associated with it. This action uses the `workspaceDelete` GraphQL mutation to permanently remove a workspace. Use this action when you need to permanently delete a workspace that is no longer needed. This action is irreversible β once deleted, the workspace and all its associated projects, environments, and data cannot be recovered.
Build the agent's code inside its own VM and deploy it. Creates the named service (default: the agent's name) or reuses it. Needs an agent booted with the builder.
Deploys all connected triggers for an environment.
Deploys a GitHub repo
Start a Railway deployment from a service's already connected repository or image source, optionally at a repository commit SHA, and return the deployment ID. This action cannot access or upload local files; connect the source to the service before calling it.
Deploy a service instance
Deploy a complete Railway template configuration into a new project or an existing project environment. Railway commits the template changes and starts their deployments.
Deprecated by Railway: Deprecated in favor of templateDeployV2 Deploys a template.
Destroy a Railway sandbox by ID in an environment. This is destructive: the sandbox can no longer run commands and requires confirm=true. Returns nullable sandbox lifecycle metadata; the returned status may reflect destruction still in progress.
Detach a project from an access group. This is destructive and requires confirm=true.
Disables point-in-time recovery on an HA database cluster with the same rolling rollout as enable. The backup bucket is left intact, so existing backup history is preserved.
Disables CDN for a service, soft-deleting the edge config.
Disconnects the authenticated user's Railway account from Discord. Use when you need to unlink a Discord account that was previously connected to Railway for notifications, authentication, or team management. This action is irreversible for the disconnect operation β the user must manually reconnect through the Railway dashboard or Discord OAuth flow if needed.
Disconnect a service from a repo This is destructive and requires confirm=true.
Dismiss a pending database security update and cancel its scheduled redeploy.
Ejects a service from the template and creates a new repo in the provided org.
Enables point-in-time recovery on an HA database cluster (Postgres HA, MySQL HA) with a rolling, near-zero-downtime rollout across its members. Safe to retry.
Enables CDN for a service, creating an edge config and attaching all live domains.
Estimate disk space needed for a Railway volume point-in-time restore at a target timestamp, optionally using a specific backup history. Returns base backup label, data/WAL/scratch sizes in MB, plan maximum size and a likely-to-fit Boolean. Returns estimate:null when unavailable. This is a conservative estimate, not a recovery guarantee. Read-only: does not restore, provision or change billing.
Execute a command inside a running Railway sandbox in the specified environment. Commands can modify the sandbox. Returns stdout, stderr, exitCode, timedOut and truncated; inspect these fields because successful API execution does not imply a successful or complete command. Do not use to access secrets or raw environment-variable values.
Force delete a scheduled deletion of a project (skips the grace period) This is destructive and requires confirm=true.
Duplicate a running cloud agent.
Generate a template for a project
Get unified AI usage for a Railway workspace, including total usage in cents, billing period end, available soft/hard limits and remaining usage. This is a read-only billing summary: it does not change spending limits or payments.
Returns the platform feature flags enabled for the current user
Get metadata for a Railway audit log entry by entry ID and workspace ID: event type, creation time and associated workspace/project/environment identifiers. Returns metadata only; unstructured event payload and context are not included. Audit logs are available to workspace admins.
Get object count and storage size for a Railway bucket instance.
Return a bounded, optionally filtered time window of build-phase logs for one Railway deployment.
Gets the image URL for a Notion image block
Get a Railway cloud agent by ID within an environment, including its lifecycle, network endpoints, repository source, and session work status. Returns cloudAgent: null when the agent belongs to another environment. Does not wake or modify the agent.
Get a Railway cloud agent's disk-checkpoint metadata by checkpoint ID, including capture status, failure reason, creation time, region and project/environment restrictions. A null environmentId permits booting anywhere in the project. Poll this lookup after checkpoint creation until status is SUCCEEDED or FAILED. Reads metadata only; does not create, restore or download the checkpoint.
Read Railway cloud-agent task metadata without waking the VM: the latest turn in a session, or an earlier task_id. Returns task/session identity, status, model settings, completion time, progress counters, active tool names and pending interaction identifiers/actions. Excludes task text, result content, attachments, JSON metadata, form fields and tool arguments. Does not submit or respond to a task.
Get compliance agreements for a workspace including HIPAA BAA and GDPR DPA status.
Read one Railway custom domain by domain ID and project ID using customDomain.status. Returns identity, routing port, synchronization, verification state, DNS record metadata and public certificate status, including certificateRetryable. Does not retry issuance or modify DNS. Verification tokens, DNS record values, private keys and free-text certificate errors are excluded.
Checks if a custom domain is available.
Progress of the HA password rotation running for a cluster, or null when none is.
Return one deployment's status, timestamps, metadata, and rollback eligibility.
Retrieve runtime logs for a deployment. Use when you need to view the runtime application logs for troubleshooting runtime issues, monitoring application behavior, or debugging errors that occur during deployment execution. This action queries the Railway GraphQL API for runtime log entries associated with the specified deployment ID. The logs are returned in reverse chronological order (most recent first).
The cloud agent `ssh dev.new` lands you in, or null when the next connection will create one.
Fetch individual DNS query logs for a Railway environment, optionally filtering by service or deployment-instance/sandbox ID and DNS log-query expression. Use anchor and before/after dates and limits to select a window. Returns queried name, record type, response code, answer IPs, ordered CNAME chain, timestamp, internal/external zone and resource identifiers. Returns one window without a pagination cursor; does not perform DNS queries or return aggregated metrics.
Deprecated by Railway: Use the `status` field within the `domain` query instead Domain with status
Get details of a specific Railway environment by its ID, including service instances and deployment information. Use this action when you need to retrieve detailed information about an environment, such as its name, creation timestamp, and the service instances deployed within it, along with their latest deployment statuses. This is a read-only action that queries the Railway GraphQL API.
Get the status or completed result metadata of an asynchronous RailwayChangeSet apply by environment ID and apply ID. Returns status, related resource IDs and per-operation kind/path/status. Omits diagnostics, operation output blobs and free-form summaries. Read-only: does not apply or modify configuration. Call again with the same IDs to check progress; performs one lookup per call.
Fetch a window of Railway project-environment logs, with optional anchor, before/after date boundaries, separate before/after limits and a log-filter query. Build logs are excluded unless the filter explicitly includes a snapshot ID. Returns timestamp, message, severity, structured attributes and resource tags. This is a single log window, not a live stream; the provider returns no pagination cursor.
Get a Railway environment patch by ID, returning lifecycle metadata: status, environment ID, timestamps, message and last application error. Does not return the configuration patch or raw variable values, and does not apply, commit or modify the patch.
Get metadata for the latest staged commit in a Railway environment: patch ID, lifecycle status, timestamps and environment identity/name. Metadata only: does not return the patch JSON, variable values, change details or free-text errors/messages, and cannot enumerate changed variable names from the opaque patch. Does not commit or apply changes.
Read Railway's end-of-current-billing-cycle usage estimates for requested measurements, optionally selecting a project or workspace and including deleted projects. Returns estimated values per project and measurement, not an invoice or guaranteed charge. This schema accepts no date range; does not change billing settings, subscriptions or payments.
Get information about a specific Railway function runtime, including its image, latest version and available version images/tags. Currently bun is the only supported runtime name. Read-only; does not create or deploy a function.
Checks if user has access to GitHub repository
Check if a repo name is available
Get information for a GitHub pull request associated with a Railway service, including its title, state, body, author, change counts, and mergeability. A null result or certain errors indicating no linked GitHub repository return empty fields.
Get info for a GitHub pull request, including fetch errors
Checks if user has access to GitHub repository
Get branches for a GitHub repo that the authenticated user has access to
Return a bounded date-anchor window of HTTP request logs for one deployment, including request, response, timing, and routing fields.
Return either HTTP request counts grouped by status code or p50/p90/p95/p99 response-time series for one Railway service and environment. Status counts are returned as evidence and are not converted into an error percentage.
Read HTTP request-count time-series samples for a Railway service and environment over an explicit start/end period, using query.httpMetrics. Optionally filter by HTTP method, request path or status code and set sample frequency in seconds. Returns Unix-second timestamps and request counts, not request logs, duration percentiles or error-rate breakdowns.
The point-in-time window a MySQL service's binlog archive can restore to, resolved from the archive bucket: the oldest full backup (floor), the newest shipped binlog (ceiling) and every full backup's timestamp. archiveConfigured is false when the service carries no complete BINLOG_ARCHIVE_* contract.
Fetch individual Railway network flow logs for an environment around an optional time anchor. Bound logs before and after the anchor with dates and limits; optionally filter by service, deployment instance/sandbox, or Railway flow expression. Returns typed traffic metadata: addresses, ports, protocols, peers, counts, latency, capture times and drop reason. Does not retrieve packet payloads or change network policy. Returns one requested window; the provider exposes no cursor or pageInfo.
Gets a notification delivery by ID for the authenticated user
Live replication health for an HA database cluster (Postgres HA, MySQL HA). Use it to gate the Enable/Disable PITR buttons so a rollout can't start while a member is too far behind to rejoin. Returns null when the service isn't an HA root.
One-shot read of the current progress of a PITR enable/disable rollout on an HA database cluster. Use it to rehydrate state on page load, then poll this field for updates. Returns null when no rollout has run recently.
Get Railway's current platform status, including its stability flag and any reported incident or maintenance details. This is platform-wide status, not the health of a particular service or deployment. Requires no input parameters.
Deprecated by Railway: Plugins are deprecated Get a plugin by ID.
Deprecated by Railway: Plugins are deprecated Fetch logs for a plugin
Get the email preferences for a user
Get a Railway service instance's private-network endpoint using environment, private network and service IDs. Returns DNS names, private IP addresses, endpoint and service-instance IDs, synchronization status, tags and nullable lifecycle timestamps. The endpoint result may be null. This is a read-only lookup; it does not create endpoints, rename DNS records or modify private networking.
Get comprehensive compliance information for a project including 2FA status, member permissions, backup schedules, and compliance agreements. Requires workspace API token with admin access.
Get resource access rules for project-specific actions
Get one page of usage grouped by project and service for a Railway workspace over a date-time interval and selected measurements. Optionally include deleted projects. first limits projects, not metric rows, to at most 50. To continue, pass page.pageInfo.endCursor as after with the same workspace, interval and measurements while page.pageInfo.hasNextPage is true. Does not modify billing or payment settings.
Get workspace members for a project with 2FA details
Fetch multiple projects by id. Skips ids the caller cannot access (does not throw on partial denial). Intended for batched dashboard hydration of a small viewport-sized set of cards.
Get public Railway stats.
Get a Railway domain by ID
Get a Railway domain by its domain name within a workspace
Whether email forwarding can be offered on this domain, or why it can't.
Caps the forwarding UI surfaces before a rule is rejected server-side. Pass workspaceId for that workspace's alias allowance; without it the highest plan's is returned.
Gets the ReferralInfo for the authenticated user.
Get resource access for the current user or workspace
Return sampled CPU, memory, disk, network, backup, or volume measurements for selected Railway resources and time range.
Get a Railway sandbox by ID in an environment, including lifecycle status, region, network isolation, idle timeout and public/private domains. Returns sandbox: null when the lookup returns no sandbox. Does not execute commands or modify the sandbox.
Get a template's build state by id (the recipe content hash). READY once its checkpoint exists; BUILDING/FAILED reflect the build workflow.
Get a Railway service by ID, including its name, owning project, creation and update timestamps, icon, and source template service ID. Use List Services to discover service IDs.
Get auto-deploy status for a Railway service instance identified by project, environment and service IDs. Returns enabled (current setting), canEnable (enablement eligibility) and an optional provider reason. Reads both flags independently; does not enable or disable auto-deploy or start a deployment.
Get explicit resource-limit overrides for a Railway service in an environment. Returns override:null when no overrides are set, rather than the effective limits merged with plan defaults. Exposes container CPU, memory bytes and process limit metadata when present. Read-only; does not allocate resources or change limits.
Read merged resource limits for a Railway service instance, including plan defaults. Returns container CPU, memory in bytes and process-ID limits when reported. These are configured limits, not current utilization measurements. Does not modify limits, change plans or expose an arbitrary provider JSON blob.
Fetch a single signal by owner scope and name.
Recent audit log entries for a signal.
Evaluate a signal against an evaluation context (pure resolution; does not mutate registry state).
Deprecated by Railway: Teams are now workspaces. Use the workspace query instead. Find a team by ID
Deprecated by Railway: Use templates instead - teams are now workspaces Get all templates for a team.
Return a Railway template's ID, name, and serialized configuration required for deterministic deployment.
Get the metrics for a template.
Get the source template for a project.
Get explicit metadata for spans of a Railway trace, oldest first, restricted to the specified environment. Includes span/parent IDs, service and deployment IDs, names, timing, component, kind and status code. Returns at most max_spans (default 1000, maximum 2000), with no continuation cursor. Omits arbitrary attributes, event/link payloads and free-form status messages. Read-only.
Read per-service tracing activity for a Railway environment within tracing retention. Returns separate timestamps for spans exported by the edge and by each service's app. Edge spans do not establish app instrumentation. A service absent from the list has produced no spans within retention; nullable timestamps mean no spans for that exporting hop in the window. Does not retrieve span payloads or enable tracing.
Get aggregated Railway usage for selected measurements across a project, workspace, or the current user's projects. Optionally choose a time range, grouping dimensions, and inclusion of deleted projects. Without start_date, Railway uses the project owner's current billing period. Values use the selected measurement's units; this does not calculate an invoice or modify billing settings. Returns the full aggregate list without pagination.
Get the authenticated Railway user's total kickback earnings. The amount may be null when earnings data is unavailable or the earnings query reports a processing problem.
Get the public profile for a user
Deprecated by Railway: Users don't have personal templates anymore, they belong to their team now Get all templates for the current user.
Get information about the user's Vercel accounts
Get metadata for a Railway volume instance by instance ID, not its parent volume ID. Returns lifecycle state, deletion flag, timestamps, sizes in MB, mount path, region and parent volume/environment/service identifiers. Service association and state may be null. Reads metadata only; does not read files, mount, resize, back up or modify the volume.
Get the status of a Railway workflow by workflow ID, including an optional provider error. States are Complete, Error, NotFound and Running. NotFound is returned as workflow state, not an HTTP failure. This read-only lookup does not start, cancel or retry workflows. Call again with the same ID to check progress.
Get a Railway workspace by ID, including its name, plan, timestamps, preferred region, and security settings. OAuth access is limited to workspaces selected during consent. Use LIST_PROJECTS to retrieve projects in the workspace.
Get the policies for a workspace
Get workspace-wide sums of Railway's UI usage rollup for selected measurements and an optional time period. Optionally include deleted projects that still bill. Returns aggregate measurement values rather than per-project rows. Read-only; does not change billing or payments. No pagination.
Details for the X build claim page: the build, whether it can be claimed, and whether the caller's connected X account owns it.
Create services and volumes from docker compose
Initiate the transfer of project ownership
Invite a user by email to a project
Leave project as currently authenticated user This is destructive and requires confirm=true.
Deprecated by Railway: Use featureFlagRemove with PRIORITY_BOARDING. Unsubscribe from the Beta program. This is destructive and requires confirm=true.
Leave a workspace This is destructive and requires confirm=true.
Get all volume instances for a given volume
Retrieve API token metadata for the authenticated Railway user, including token IDs, names, and creation timestamps. A null token result is returned as an empty list.
List all Railway audit log event type identifiers and their descriptions. Returns the event type catalog, not actual audit log records or user activity. Accepts no workspace filters or pagination.
List one cursor-paginated page of Railway workspace audit-log metadata. Requires workspace admin access. Optionally filter by time, event types, project or environment and choose sort order. Returns event IDs, types, timestamps, resource IDs and pageInfo; raw payload/context and actor details are excluded. Continue forward with endCursor as after or backward with startCursor as before, preserving filters and sort. Retention depends on the workspace plan.
Railway agent chat bindings (Slack/Discord scope β workspace) across the caller's workspaces.
List one cursor-paginated page of Railway cloud-agent disk checkpoints in an environment, newest first. Returns checkpoint identity, capture status, region, optional failure reason and environment boot restriction, not disk contents. Continue forward with pageInfo.endCursor as after or backward with pageInfo.startCursor as before, keeping the same environment_id. Does not create checkpoints, restore disks or modify agents.
List one page of resumable shell and exec sessions on a Railway cloud agent's VM. Returns session names, kinds, attachment/foreground state, process lifecycle and native pageInfo cursors. Returns sessions:null if the agent has no running machine. Does not wake the machine, attach to sessions or execute commands. Command text and connection credentials are excluded.
List the models Railway cloud-agent tasks can run on in a project environment, with provider/display names, supported reasoning efforts and the default model and reasoning effort for new conversations. Does not dispatch a task or change defaults.
List one page of Railway cloud-agent task records in an environment, newest first. Optionally filter by agent, task status, or one session's turns (session_id requires cloud_agent_id). Returns typed lifecycle timestamps, IDs, model settings and correlation metadata. Excludes prompts, task output, error content, attachments, and JSON blobs. Continue with nextCursor as cursor, preserving environment and filters. Does not submit or modify tasks.
List persistent coding cloud agents in an environment, including lifecycle and work status, networking, and repository source. Optionally filter by archive state or ownership. Returns the complete list without pagination.
List one cursor-paginated page of Railway deployment-instance executions for a service in an environment. Returns execution/deployment IDs, lifecycle status, creation/update timestamps and nullable completion time. Excludes opaque deployment configuration metadata. Continue forward with pageInfo.endCursor as after or backward with pageInfo.startCursor as before, retaining service_id and environment_id. This is a read-only listing, not command execution.
List one page of deployment triggers for a Railway service in a project and environment. Returns repository, branch, provider and check-suite configuration. Use first/after for forward pagination or last/before for backward pagination with returned edge and pageInfo cursors. Does not trigger a deployment or modify configuration.
List one page of Railway deployments for a project, environment, or service, optionally filtered by status or deletion state.
List Railway and custom domains for a service, including IDs, ports, DNS records, certificate state, and custom-domain verification token.
List all egress gateways assigned to a Railway service instance identified by service and environment IDs. Returns gateway IPv4 addresses, regions and nullable zones. This is a read-only list: it does not allocate or assign gateways or change networking. The provider returns no pagination cursor.
List one cursor-paginated page of settled Railway environment history, newest first: system events and terminal operations. Filter by action, object, outcome, services, time, or dashboard activity-feed mode. Returns actor and change metadata without configuration payloads or variable values. Continue with pageInfo.endCursor as after when hasNextPage is true.
Retrieve a page of patches for a Railway environment, including identifiers and timestamps. The request accepts a page size and an after cursor, but the response does not return pagination information.
List pending project-mutating work for a Railway environment, combining patches and standalone deploy operations. Returns parent operations with direct in-flight child metadata, statuses, timestamps, references, workflow IDs and actor metadata. Configuration changes and variable values are excluded at every level; deeper descendants are not expanded. Returns the full list without pagination. This live operation view is not a stored inbox or settled history. Does not apply, cancel or change pending work.
List one page of project environments with IDs, pull-request metadata, source environment, and staged-change count.
Get the workspaces the user doesn't belong to, but needs access (like when invited to a project)
List available Railway function runtimes, including each runtime's image, latest version and available version image tags. Returns the runtime catalog; does not create, deploy or modify a function. Accepts no arguments or pagination.
Retrieve a list of GitHub repositories that Railway has access to. Use this action when you need to fetch the list of GitHub repositories associated with the authenticated user's GitHub account through Railway. This is a read-only operation that returns the id, name, and owner for each repository. The action queries the Railway GraphQL API githubRepos endpoint and returns the complete list of accessible GitHub repositories.
Retrieve the list of GitHub scopes the user has installed the installation to. Use this action when you need to fetch the list of GitHub scope names that the authenticated user has granted to the Railway GitHub application. This is a read-only operation that returns an array of GitHub permission scope strings such as 'repo', 'workflow', or 'read:user'. Note: This endpoint requires account-level authentication, not workspace/project tokens.
Get the Herokus apps for the current user
List Railway integration authorizations associated with the authenticated user. This read-only operation returns integration IDs, provider names and provider IDs.
Get all integrations for a project
Find cloud agents you own across every Railway project and environment you can access in one request. Optionally filter by archive state; omitted or null includes both archived and unarchived agents. Returns identifying metadata, last observed lifecycle state and public/private networking metadata, without session contents or credentials. Not paginated. Use List Cloud Agents to list an environment's agents.
Retrieve notification deliveries for the authenticated user, including their IDs, types, and creation timestamps.
Get all notification rules for a workspace and project
List one cursor-paginated page of Railway observability dashboards in an environment. Returns dashboard/item IDs, panel names, descriptions, types, selected metric/resource metadata and attached monitor identities/timestamps. Excludes layout blobs, free-form log filters, monitor configuration and alert content. Continue forward with pageInfo.endCursor as after or backward with pageInfo.startCursor as before, retaining environment_id. Does not run dashboard queries or modify dashboards.
List private networks for a Railway environment, including DNS name, network name, numeric and public identifiers, project/environment IDs, tags and nullable creation/deletion timestamps. Returns one unpaginated list. Does not create, modify, connect to or delete a private network.
Get one page of Railway project events, optionally filtering by environment, action, object type or service ID. Returns event metadata and pageInfo for forward/backward cursor pagination. Does not return untyped event payloads or activity payloads, which may contain sensitive data. Read-only.
The authenticated user's favorited project ids in a workspace, in the order the Favorites row should render them. Defaults to the order they were starred in.
Get invitations for a project
List users belonging to a Railway project with their user IDs, email addresses, project roles, and available names and avatars. This read-only endpoint has no pagination. Does not invite users or change membership.
List one page of projects in an accessible Railway workspace with project and workspace identifiers. Use LIST_SERVICES and LIST_ENVIRONMENTS for those independently paginated resources.
List DNS records for a Railway domain
List email forwarding rules for a Railway domain
Get Railway domains for a workspace
Fetch trusted domains configured for Railway workspace membership invitations, optionally specifying a workspace ID. The results can be used to audit domain configurations or obtain domain IDs.
List available Railway regions, optionally passing a project ID. Returns country, location, identifiers and deployment constraints, including deprecation and replacement-region details when available. Returns a single list without pagination.
List named sandbox checkpoints in an environment.
Resumable sessions inside a sandbox β interactive shells and one-off exec commands, live or recently exited. Null when the sandbox can't report them (e.g. its vm-init predates session listing).
List sandboxes in a Railway environment, newest first, with lifecycle state, region, idle timeout, network isolation and published domains. Filter by active or destroyed state. Without pagination arguments returns the full list; use first/after for forward pages or last/before for backward pages. Returns edge cursors and pageInfo for continuation.
List one page of service IDs and names in a project for service, deployment, logs, metrics, domain, and volume tools.
List signals registered for an owner scope.
List all Railway TCP proxies for a service in an environment, including public hostnames and ports, target application ports, synchronization states and available timestamps. Returns an empty list when no proxies exist. Does not create or modify network exposure. This endpoint has no pagination.
List one page of all published Railway templates, optionally restricted to recommended or verified templates. Uses query.templates, without a text search. Set first with after for forward pagination or last with before for backward pagination, retaining filters between calls. Returns summary metadata and native Relay cursors.
List Railway environment traces with at least one span matching the filter, newest first. Optionally restrict by service and span-start time range; the default range is the last hour, clamped to project retention. Returns typed trace summaries, timing, error/span counts and root-service metadata, without raw span or resource attribute blobs. limit defaults to 100, maximum 500. This is a single bounded result list; the provider supplies no pagination cursor.
Get all trusted domains for a workspace
Return Railway shared-environment or service variables. Choose stored to preserve or render references, or deployment_resolved to return the values supplied to a specific service deployment. Variable values may be sensitive.
List backup schedules for a Railway volume instance, including names, IDs, frequency categories, cron expressions, creation times and optional retention durations in seconds. Returns schedules, not backup snapshots or database contents. Does not create, change or run a schedule.
List backups of a Railway volume instance, returning backup IDs, names, creation and optional expiry timestamps, creator/schedule identifiers and available size metadata in MB. Use a volume instance ID, not the project-level volume ID. Returns the complete provider list without pagination; an empty list is valid. Reads only backup metadata, not backed-up contents. Does not create, restore or delete backups.
List one page of persistent volume IDs and names in a project for storage configuration and metrics workflows.
Gets all identity providers of a workspace
List deploy sources that can be added to a workspace policy.
Get all templates for a workspace.
Marks notification deliveries as read
Merge a canvas layout from one environment into another. Re-computes the merge from current state and applies mutations.
Apply a pending database security update now (backup, then redeploy). Returns the new deployment id.
Preview the HA static egress IPs that would be assigned, without applying them. Set allEnvironments to cover all of the service's environments.
Preview the high-availability static egress IPv4 addresses that Railway would assign to a service in an environment, including region and optional availability zone. This read-only preview does not persist assignments or enable/configure static egress. Returns the provider's full preview list without pagination.
Preview the legacy static egress IPv4 addresses that Railway would assign to a service in an environment, including region and optional zone. This read does not persist an assignment, enable static outbound networking or reserve an IP.
Publish or update an existing Railway template in the public template marketplace, where other users can view and deploy it. Set category, description and README, optionally an image, demo project and workspace. A demo project becomes accessible through the public Live Demo link. Publishes the template's existing configuration; ensure that configuration and supplied public text contain no secrets or raw environment-variable values. Returns metadata only, not template configuration.
Recover the latest task response from the agent. May wake the VM.
Create a new Railway deployment with the selected deployment's exact code and build/deploy configuration, optionally using its previous image tag. This does not deploy the latest source commit and requires confirm=true.
Redeploy a service instance
Remove a member from an access group. This is destructive and requires confirm=true.
Remove a Railway agent chat binding. Requires admin on the binding's workspace. This is destructive and requires confirm=true.
Remove a feature flag for a user This is destructive and requires confirm=true.
Remove a feature flag for a project This is destructive and requires confirm=true.
Remove user from a project This is destructive and requires confirm=true.
Stop forwarding every address on this domain, and remove the mail records that delivered them This is destructive and requires confirm=true.
Remove a feature flag for a service This is destructive and requires confirm=true.
Deprecated by Railway: Use featureFlagRemove with PRIORITY_BOARDING. BETA leaves Priority Boarding. Remove a flag on the user. This is destructive and requires confirm=true.
Remove a deploy source from a workspace policy allowlist. This is destructive and requires confirm=true.
Remove a user from a workspace This is destructive and requires confirm=true.
Set a session's title, shown over its generated one. An empty title clears it.
Rename one environment without changing or deploying its configuration.
Rename a private network endpoint.
Rename a sandbox checkpoint.
Replace a signal's type and default, clearing all rules. Destructive β intended for CLI --force re-type.
Replace the entire Railway shared-environment or service variable collection with the supplied complete map. This is irreversible: every existing variable whose key is omitted from the map is permanently deleted. Railway skips deployments by default; set skip_deploys=false and confirm_deploys=true to explicitly trigger deployments.
Report the current coding-agent session state.
Resend an invitation for a project
Reset an observability dashboard to default dashboard items
Deprecated by Railway: Plugins are deprecated Reset envs and container for a plugin in an environment
Answer a question the task is waiting on. Wakes a sleeping VM to deliver it (the question survives sleep); poll cloudAgentTask for the outcome.
Copy a FAILED patch's changes into the environment's staged patch (creating one if needed). The FAILED patch is left untouched.
Restart a running or crashed Railway deployment by reusing its existing image without rebuilding or creating a deployment from newer source. Requires confirm=true.
Deprecated by Railway: Plugins are deprecated Restarts a plugin.
Start restoring a Railway volume instance from a backup. This overwrites target volume data. Optionally restore HA replica services and wipe specified services' volumes to fresh empty volumes. Returns asynchronous workflow metadata, not proof that restoration has finished.
Start a point-in-time restore from a PostgreSQL or MySQL database volume instance's PITR backup history. Creates and deploys a brand-new sibling database service and volume in the project; the source stays online and untouched. The new service inherits source environment configuration and recovery references. Requires an existing base backup and archived logs covering the target. Returns nullable workflow metadata, not restored data or completion status; cutover is separate.
Retrigger verification for a failed trusted domain
Read the corrected Railway domainStatus object, then retry custom-domain certificate issuance only when Railway reports certificateRetryable=true. A non-retryable result fails without mutation and includes status evidence.
Reverts an HA cluster to standalone mode using template metadata to derive variables to remove
Restore a previous Railway deployment's Docker image and custom variables as the active deployment without rebuilding. Railway must report canRollback=true; the operation requires confirm=true.
Restore a signal to the snapshot captured before a change (reads prevState from the target change row).
Search one cursor-controlled page of Railway templates by text, verification state, and category.
Send a message to a forwarded address so the customer can confirm it delivers.
Extend a sandbox's lifetime.
Make this cloud agent the box `ssh dev.new` lands you in. Replaces any previous choice.
Star or unstar a project for the authenticated user. Returns the resulting state.
Delegate the domain's authoritative nameservers to an external DNS provider, or reset to Railway defaults by passing an empty list.
Enables or disables Under Attack Mode for a service. While enabled, the edge serves a browser challenge to unverified visitors of the service's edge-routed domains; non-browser clients (APIs, webhooks) without a clearance cookie are rejected with a 429. Optionally time-boxed via durationSeconds, after which the mode disarms automatically.
Change a signal's canonical default (production floor).
Attach or replace a rule on a signal. Matching rules must agree at resolution time or the default is returned.
Deprecated by Railway: Use featureFlagAdd with PRIORITY_BOARDING. BETA joins Priority Boarding. Set flags on the authenticated user.
Sleep a running cloud agent, keeping its volume. Processes on the machine are terminated; waking re-runs its entrypoint.
Skip the next scheduled auto-update for a service instance. Takes effect immediately, with no staged change and no redeploy. The maintenance window and update policy stay as they are. This only delays the next scheduled update.
Create all resources from a complete Railway template configuration in a new project or an existing project environment, but leave the resulting changes staged. This does not commit the staged changes or start deployments; commit them separately when ready.
Deprecated by Railway: Plugins are deprecated Force start a plugin
Stop a currently running Railway deployment without deleting the service or its deployment-history record. Requires confirm=true.
Transfer resources between named IaC partials without deleting, redeploying, or changing them. Omit resources to transfer the entire source partial. The destination may be a new or existing partial.
Transfer a project to a workspace
Bring an archived cloud agent back. Owner only.
Unpublish a Railway template by ID. Requires permission to manage the template. Returns Railway's Boolean acknowledgement as unpublished; a false value does not confirm unpublishing. Does not delete the template.
Update an access group's name or role.
Sets the base environment override for a deployment trigger.
Rename an existing Railway bucket by ID and return its bucket metadata. This updates the display name without changing environment configuration.
Record that the agent's owner looked, so its DONE and ERROR outcomes stop being unseen. Without sessionId, every session. Owner only.
Start a task and return its handle. Creates a fresh agent by default; replyToTaskId continues the original conversation.
Updates a deployment trigger.
Update a generated Railway domain's hostname and optionally its target port, or update a custom domain's target port. Generated-domain updates require the service ID and hostname.
Switch a service from HA static egress IPs back to a standard static IP. Set allEnvironments to apply to all of the service's environments.
Sets whether Railway Authentication protects an environment's services that have no setting of their own, and pushes it to them. Requires a workspace admin; turning it on also requires the workspace's Railway Authentication entitlement. Services with edge caching on that it protects have their caching turned off.
Release resources from a named IaC partial without deleting, redeploying, or changing them. Omit resources to release the entire partial. Whole-project management is possible once no named partials remain. A later named-partial apply can claim released resources again.
Deploy a service that was skipped when its environment was created, along with its skipped transitive dependencies.
Agree to the fair use policy for the currently authenticated user
Updates a GitHub repo through the linked template
Update an integration for a project
Update a notification rule
Update an observability dashboard
Deprecated by Railway: Plugins are deprecated Updates an existing plugin.
Update the email preferences for a user
Update project settings and configuration on Railway. Use this action when you need to modify project properties such as description, visibility settings, or PR deploy preferences. The following fields can be updated: - description: A brief description of the project - isPublic: Whether the project is publicly visible - prDeploys: Whether pull request preview deployments are enabled This is an idempotent operation β updating a project with the same settings multiple times will not change the result after the first update.
Sets whether environments created from now on start with Railway Authentication protecting services that have no setting of their own. Existing environments are unchanged. Requires a workspace admin; turning it on also requires the workspace's Railway Authentication entitlement.
Change the role for a user within a project
Deprecated by Railway: Use projectTransfer instead Transfer a project to a team
Purges the CDN cache for a service. Bumps the edge config's purge epoch so every edge node treats prior cached entries as stale on next request. Idempotent; returns true even if CDN is disabled for the service.
Update a Railway domain's settings
Update a DNS record for a Railway domain
Change where a forwarded address delivers
Updates the ReferralInfo for the authenticated user.
Update a service's name or icon without changing environment-specific deployment settings.
Updates the edge config (caching settings) for a service.
Replaces the service's edge rules (whole-section replace; null clears). Rules are boolean expressions over request attributes plus one action (block/allow/challenge/redirect/cache_override), evaluated at the edge before requests reach the service. The ruleset is validated server-side; invalid rulesets are rejected with per-rule diagnostics and nothing is written. Rules without an id get a server-assigned stable `rul_` id.
Update build/deploy settings for a service in a specific environment on Railway. Use this action when you need to modify service configuration including commands, healthcheck, replicas, region, and cron schedule. This is an idempotent operation. The following fields can be updated: - startCommand: Custom start command for the service - buildCommand: Custom build command - rootDirectory: Root directory for monorepo setups - healthcheckPath: Health check endpoint path - healthcheckTimeout: Health check timeout in seconds - region: Deployment region - numReplicas: Number of service replicas - restartPolicyType: Restart policy (ON_FAILURE, ALWAYS, NEVER) - restartPolicyMaxRetries: Maximum restart retries - cronSchedule: Cron expression for scheduled deployments - sleepApplication: Enable/disable sleep when idle - dockerfilePath: Custom Dockerfile path - watchPatterns: File patterns for automatic redeployment
Enables or disables auto-deploy for a service instance.
Update the auto-update maintenance window for a service instance, applied immediately (no config staging, no redeploy). Only the schedule changes β the update policy is untouched.
Sets a service instance's own Railway Authentication setting: on, off, or none (follow the environment's default). Requires a workspace admin; ending up on also requires the workspace's Railway Authentication entitlement and edge caching off.
Update the resource limits for a service instance
Remove the upstream URL from all service instances for this service
Configure a shared variable.
Remove a rule from a signal.
Sets the default size (in MB) for a volume mount in a template's config. New volumes created when the template is deployed are provisioned at this size, clamped to the deployer's plan maximum. Pass sizeMB: null to clear the pre-size and fall back to the plan default. Editing a template requires maintainer access.
Update the role of a trusted domain
Updates the profile for the authenticated user
Update date of TermsAgreedOn
Rename a Railway volume without changing any volume instance.
Update a Railway volume's environment-specific mount path, attached service, or state. service_id is required on every call as the current or target attachment; set it to null only to explicitly keep or make the instance disconnected.
Removes backup expiration date
Manage schedule for backups of a volume instance
Update a Railway workspace's name, avatar, or preferred region. Omitted fields stay unchanged; explicit null values are sent to Railway. Requires admin access to the workspace.
Changes a user workspace permissions.
Enable or disable a workspace policy. Enterprise workspaces only.
Control whether non-admin project visibility is restricted to access groups and direct project permissions.
Enable HA static egress IPs for a service. Set allEnvironments to apply to all of the service's environments.
Create or return a workspace Slack channel through Railway's upsertSlackChannel route, which identifies the workspace with workspace_id.
Create or update a non-empty map of Railway shared-environment or service variables without deleting variables omitted from the submitted map. Railway skips deployments by default. Set skip_deploys=false and confirm_deploys=true to explicitly trigger deployments.
Create or return a workspace Slack channel through Railway's workspaceUpsertSlackChannel route, which identifies the workspace with id.
Validates an edge ruleset without writing it β returns the same diagnostics updateServiceEdgeRules would reject with (empty list = valid). For live feedback in rule editors.
Wake a sleeping cloud agent.
Nothing matches that.
More in Developer tools
See them all βGitHub is a code hosting platform for version control and collaboration, offering Git-based repository management, issue tracking, and continuousβ¦
883 toolsβ SSupabase is an open-source backend-as-a-service providing a Postgres database, authentication, storage, and real-time subscription APIs for buildingβ¦
230 toolsβ BBitbucket is a Git-based code hosting and collaboration platform supporting private and public repositories, enabling teams to manage and review codeβ¦
116 toolsβ SIntegrate Sentry to manage your error tracking and monitoring.
210 toolsβ CCoinbase is a platform for buying, selling, transferring, and storing cryptocurrency.
28 toolsβ AAppDrag is a cloud-based platform for building websites, APIs, and databases with drag-and-drop tools, code editing, and integrated hosting toβ¦
11 toolsβ