MCP Server / guide 7 of 19
Tool access is controlled by the credential's capabilities and Actor allowlist. Rewrite and optimization tools create proposals; applying or publishing remains a separate, explicit operation.
The server registers 52 tools. Call get_mcp_capabilitiesto see which ones your credential can use. Inventory, usage, and revenue reads use synchronized data; optimization and publishing runs may continue in the background.
| Tool | Capability | Inputs | Behavior |
|---|---|---|---|
| get_workspace | workspace:read | None. | Read the workspace and Apify connection status bound to this credential. |
| get_mcp_capabilities | MCP credential | None. | List this credential's permissions, Actor scope, tools, resources, and prompts. |
| get_mcp_usage | workspace:read | days (optional, 1–90; default 30). | Read this credential's per-tool calls, failures, latency, and last use. |
| Tool | Capability | Inputs | Behavior |
|---|---|---|---|
| list_actors | actors:read | status, page, limit (1–500), search, readiness (all optional; page 1 and limit 100 by default). | Page through synchronized tracked Actors with status, search, and readiness filters. |
| get_actor | actors:read | apifyActorId (required). | Read one tracked Actor's cached metadata and recent copy and pricing history. |
| list_published_actors | actors:read | limit (optional, 1–500; default 100). | List public Actors that are neither deprecated nor under maintenance. |
| list_maintenance_actors | actors:read | limit (optional, 1–500; default 100). | List tracked Actors marked as under maintenance. |
| list_deprecated_actors | actors:read | limit (optional, 1–500; default 100). | List tracked Actors marked as deprecated. |
| list_unpublished_actors | actors:read | limit (optional, 1–500; default 100). | List tracked private Actors. |
| list_publish_ready_actors | actors:read | limit (optional, 1–500; default 100). | List Actors whose synchronized listing passes readiness checks. |
| sync_actors | actors:sync | None. | Synchronize the tracked Actor inventory from the connected Apify account. |
| Tool | Capability | Inputs | Behavior |
|---|---|---|---|
| list_notifications | notifications:read | unreadOnly, source, type, from/to ISO timestamps, page (1–100), limit (1–100); all optional. | Page and filter workspace and available Apify notifications. |
| get_unread_notification_count | notifications:read | None. | Count unread workspace and available Apify notifications. |
| mark_notification_read | notifications:write | notificationId (required). | Mark one workspace or Apify notification as read. |
| mark_all_notifications_read | notifications:write | None. | Mark local notifications read and attempt the same for Apify notifications. |
| Tool | Capability | Inputs | Behavior |
|---|---|---|---|
| analyze_actor | actors:optimize | apifyActorId (required). | Queue an Actor listing audit with reviewable improvement proposals. |
| rewrite_actor_title_description | actors:optimize | apifyActorId (required). | Queue reviewable title and description proposals. |
| rewrite_actor_readme | actors:optimize | apifyActorId (required). | Queue a reviewable README proposal from synchronized source. |
| rewrite_actor_listing | actors:optimize | apifyActorId (required). | Queue reviewable title, description, and README proposals. |
| generate_actor_schemas | actors:optimize | apifyActorId (required). | Queue reviewable input, output, and key-value-store schema proposals. |
| create_actor_optimization_run | actors:optimize | actorIds (1–10,000 ActorConsole or Apify Actor IDs) and fields (required); kind (optional, default CUSTOM). | Queue proposal generation for selected Actors and fields. |
| get_actor_optimization_run | actors:optimize | runId (required). | Read a proposal run's results, scores, issues, and status. |
| apply_actor_changes | actors:write | runId, items (resultId plus selected fields), and idempotencyKey (all required). | Approve selected proposal fields and queue application to Apify. |
| get_actor_change_status | actors:optimize | runId (required). | Read the same proposal run status returned by get_actor_optimization_run. |
| redesign_readme_from_files | actors:optimize | title and files (required); description and keywords (optional). Up to 40 files, 100,000 characters each, 500,000 combined. | Generate a README and patch from file contents supplied by the client. |
| Tool | Capability | Inputs | Behavior |
|---|---|---|---|
| get_pricing | pricing:read | apifyActorId (required). | Read one Actor's synchronized current pay-per-event pricing. |
| set_pricing | pricing:write | apifyActorId and pricing (required); reason (optional). | Apply pay-per-event pricing and record a change when Apify accepts it. |
| recommend_pricing | pricing:read | apifyActorId (required). | Compare live Store listings and return a median-price recommendation. |
| get_pricing_history | pricing:read | apifyActorId (required); limit (optional, 1–180; default 60). | Read up to 180 recorded pricing changes, newest first. |
| search_actors | actors:read | query (required); limit (optional, 1–50; default 20). | Search the synchronized public Store catalog and return results with demand metrics. |
| keyword_competitors | actors:read | keyword (required); limit (optional, 1–50; default 20). | Return public Store matches and search metrics for a keyword. |
| Tool | Capability | Inputs | Behavior |
|---|---|---|---|
| get_usage_costs | usage:read | apifyActorId (required); days (1–365, default 30), granularity, and mode (optional). | Read one Actor's synchronized runs and platform costs by day or month. |
| get_actor_performance | usage:read | apifyActorId (required); limit (optional, 1–180 days; default 60). | Read one Actor's cached run, result, success, and platform-cost metrics. |
| get_actor_revenue | usage:read | apifyActorId (required); limit (1–180 days, default 90), granularity, and mode (optional). | Read one Actor's creator revenue, costs, profit, coverage, and report points. |
| get_workspace_usage | usage:read | granularity (day/month) and mode (absolute/cumulative); both optional. | Read workspace usage, costs, runs, and per-Actor data. |
| get_workspace_revenue | usage:read | granularity (day/month) and mode (absolute/cumulative); both optional. | Read workspace revenue, cost, profit, margin, and Actor data. |
| get_usage_history | usage:read | granularity (day/month) and mode (absolute/cumulative); both optional. | Read the combined workspace usage and revenue history. |
| Tool | Capability | Inputs | Behavior |
|---|---|---|---|
| change_actor_store_status | publishing:write + actors:write | actorIds (ActorConsole IDs), action, and idempotencyKey (all required). | Queue immediate maintenance or deprecated status changes for selected Actors. |
| list_actor_status_operations | publishing:read | page (default 1) and limit (1–100, default 20); both optional. | Page through durable Store status operations. |
| get_actor_status_operation | publishing:read | operationId (required). | Read one Store status operation and its per-Actor results. |
| list_publishing_schedules | publishing:read | None. | List workspace publishing schedules and recent runs. |
| get_publishing_schedule | publishing:read | scheduleId (required). | Read one ordered publishing schedule and recent runs. |
| create_publishing_schedule | publishing:write | name, actorIds (ActorConsole IDs), and idempotencyKey (required); cronExpr, dailyLimit, and enabled (optional). | Create an ordered schedule, paused by default. |
| update_publishing_schedule | publishing:write | scheduleId and idempotencyKey (required); name, actorIds, cronExpr, dailyLimit, and enabled (optional). | Edit schedule settings or replace its ordered Actor queue. |
| add_actors_to_publishing_schedule | publishing:write | scheduleId, actorIds (ActorConsole IDs), and idempotencyKey (all required). | Add Actors to a publishing schedule queue. |
| remove_actors_from_publishing_schedule | publishing:write | scheduleId, actorIds (ActorConsole IDs), and idempotencyKey (all required). | Remove Actors from a publishing schedule queue. |
| pause_publishing_schedule | publishing:write | scheduleId and idempotencyKey (required). | Pause a publishing schedule. |
| resume_publishing_schedule | publishing:write | scheduleId and idempotencyKey (required). | Resume a publishing schedule. |
| run_publishing_schedule | publishing:run | scheduleId and idempotencyKey (required). | Queue an immediate publishing run, subject to Apify quota. |
| list_publishing_runs | publishing:read | scheduleId (required). | List recent execution records for a schedule. |
| get_publishing_run | publishing:read | scheduleId and runId (required). | Read one execution record and its per-Actor results. |
| requeue_publishing_actor | publishing:write | scheduleId, actorId (ActorConsole ID), and idempotencyKey (all required). | Move one blocked or failed Actor back to pending. |
list_actors returns pages of up to 500. Status-specific lists return the first 100 by default and accept a limit up to 500. The Actors resource exposes the first 500; use the tool to page further.
get_actor_revenue returns creator revenue and profit from the synchronized financial ledger. Its limit is a 1–180 day window; day/month and absolute/cumulative report modes are supported. get_actor_performance returns run and platform-cost data, not creator revenue.
get_pricing reads synchronized pricing. get_pricing_history returns up to 180 recorded changes. set_pricing writes pay-per-event pricing and accepts primary and one-time event flags. recommend_pricing compares live public Store listings. Keyword searches use the synchronized public Store catalog and include summary metrics.
For set_pricing, pricing.model must be PAY_PER_EVENT. pricing.events needs at least one named event. Each event requires a nonnegative eventPriceUsd and nonempty eventTitle; eventDescription, isPrimaryEvent, and isOneTimeEvent are optional. At most one event can be primary.
list_actors.status accepts all, published, maintenance, deprecated, unpublished, or publish-ready. readiness accepts all, ready, needs-attention, unknown, missing-description, missing-readme, missing-input-schema, missing-output-schema, or missing-key-value-store-schema. change_actor_store_status.action accepts mark_maintenance, clear_maintenance, mark_deprecated, or clear_deprecated.
redesign_readme_from_files accepts up to 40 files, 100,000 characters per file, and 500,000 characters combined. Hosted HTTP requests also have a 1 MiB body limit.
create_actor_optimization_run.kind accepts GENERATE_IMPROVEMENTS, REWRITE_COPY, REWRITE_README, GENERATE_OUTPUT_SCHEMA, UPDATE_PRICING, or CUSTOM. Available fields are title, description, readme, inputSchema, outputSchema, keyValueStoreSchema, and pricing. Actor IDs in this tool may be ActorConsole IDs or Apify Actor IDs.
Publishing tools take ActorConsole internal Actor IDs from list_actors items' id fields. Most single-Actor read tools take apifyActorId. Publishing mutations and apply_actor_changes require idempotency keys. Schedule creation is paused by default. A restricted credential can include up to 100 tracked Actors.
actorconsole://workspaceRequires workspace:readactorconsole://actorsRequires actors:readactorconsole://actors/{actorId}Requires actors:readactorconsole://notificationsRequires notifications:readactorconsole://publishing/schedulesRequires publishing:readoptimize_actor_listingRequires actors:optimizeredesign_actor_readmeRequires actors:optimizetriage_maintenance_actorsRequires actors:readprepare_publishing_scheduleRequires publishing:write + actors:readreview_deprecated_actorsRequires actors:read