Postman
Arcade.dev tools for interacting with Postman
0.1.0Postman is an API development and testing platform; this toolkit lets Arcade agents manage Postman workspaces, collections, environments, mock servers, monitors, and API definitions programmatically.
Capabilities
- Collection management — create, inspect, update, fork, export as OpenAPI, and permanently delete collections; add, rename, or remove folders and requests within them.
- Environment management — create, inspect, update individual variables, and delete environments.
- Mock servers — create mock servers from collections, inspect their public URLs, and delete them.
- Monitors — create scheduled health monitors, inspect their run history and schedule, trigger on-demand runs with pass/fail results, and delete them.
- API definitions & schemas — list APIs in a workspace, inspect API definitions, and read schema files with their content.
- Workspace & account inspection — list and inspect workspaces (with all their resources), list collections/environments/mocks/monitors, and verify which Postman account and plan the API key belongs to.
Secrets
POSTMAN_API_KEY — A Postman API key that authenticates every request made by this toolkit. Generate one in your Postman account under Settings → API keys (Postman API key docs). The key inherits the permissions of the generating account; for full toolkit functionality (creating/deleting resources across workspaces) the account must have the appropriate workspace role. Free and paid plans both support API keys, but certain workspace types or team features may require a paid plan.
See the Arcade secrets guide for how to store secrets, or add them directly at https://api.arcade.dev/dashboard/auth/secrets.
Available tools(30)
| Tool name | Description | Secrets | |
|---|---|---|---|
Add a folder to a collection, optionally nested inside an existing folder. | 1 | ||
Add a request to a collection, optionally inside a folder. | 1 | ||
Create a mock server from a collection so a client can call its simulated endpoints. | 1 | ||
Create a monitor that runs a collection on a schedule to watch an API's health. | 1 | ||
Permanently delete a collection. This cannot be undone. | 1 | ||
Delete a folder or request from a collection. This cannot be undone.
Deleting a folder also removes the requests it contains. | 1 | ||
Permanently delete an environment. This cannot be undone. | 1 | ||
Permanently delete a mock server. This cannot be undone. | 1 | ||
Permanently delete a monitor. This cannot be undone. | 1 | ||
Fork a collection into a workspace as an independent, editable copy. | 1 | ||
Inspect an API definition, including its name, summary, and attached schemas. | 1 | ||
Read an API schema's files and their definition content. | 1 | ||
Inspect a collection and return its variables and a flat tree of its folders and requests. | 1 | ||
Export a collection as an OpenAPI definition. | 1 | ||
Inspect an environment and return its variables. | 1 | ||
Inspect a mock server, including its public URL and the collection it is based on. | 1 | ||
Inspect a monitor, including its run schedule and most recent run result. | 1 | ||
Inspect a workspace and list the collections, environments, mocks, and monitors in it. | 1 | ||
List the API definitions in a workspace. | 1 | ||
List collections, optionally scoped to a workspace and filtered by name.
Without a workspace this returns the collections the API key can access: those you own or
have subscribed to. A collection another team member created in a shared workspace may not
appear here until it is subscribed to; use get_workspace to see everything a workspace holds. | 1 | ||
List environments, optionally scoped to a workspace and filtered by name. | 1 | ||
List mock servers, optionally scoped to a workspace and filtered by name. | 1 | ||
List monitors, optionally scoped to a workspace. | 1 | ||
List the workspaces the API key can access, optionally filtered by type. | 1 | ||
Trigger a monitor to run now and return its pass/fail results.
The run is synchronous: Postman holds the connection until the collection finishes. A run
that outlasts the tool's bounded wait returns ``timed_out=true`` while still executing
upstream; read the outcome from get_monitor's last-run fields rather than retrying. | 1 |