Skyvern
for LuumenAI
See what your Skyvern agents are doing, and stop the runs that went wrong
Skyvern runs browser agents that click through websites, fill forms, and pull data back. Luumen connects with an API key and reads the whole picture: which agents exist, how they are organized into folders, and what every task and workflow run is doing right now. When a run is stuck or misfiring, Luumen proposes the cancel and waits for your approval before sending it.
The Skyvern toolbox
10 tools: 5 read, 5 write. Reads answer instantly. Writes require approval by default. Everything is logged.
- ReadGet FolderGet one Skyvern folder by ID, including its current agent count.
- ReadGet RunGet current status, output, errors, files, and observation links for a task or workflow run.
- ReadList AgentsSearch and list the latest versions of Skyvern agents available to the connected organization.
- ReadList FoldersSearch and list folders used to organize Skyvern agents.
- ReadList RunsFind task and workflow runs for the connected organization, with status, type, workflow, and text filters.
- WriteCancel RunsCancel one or more active task or workflow runs in one request and report which IDs succeeded or failed. Approval by default
- WriteCreate FolderCreate a folder for organizing Skyvern agents. Approval by default
- WriteDelete FolderDelete a Skyvern folder without deleting the agents it contains. Approval by default
- WriteStart Browser TaskStart an autonomous browser task that can navigate and interact with websites, including clicking and submitting forms. Approval by default
- WriteUpdate FolderChange the title or description of an existing Skyvern folder. Approval by default
One prompt, start to finish
What a governed Skyvern run looks like inside Luumen.
Questions
Authorize once with API token. Luumen lists the scopes each action needs before you approve the connection, and credentials never appear in the chat.
Read actions answer immediately. Anything that writes — cancel runs, create folder, delete folder, start browser task, and more — is shown as a plan and requires approval by default, including the 3 actions classified as destructive. Administrators configure that per tool, so you decide exactly which actions can ever run unattended.
You decide. Actions are granted per agent, skill, and team, and per environment — production is not staging. Read access can be broad while writes stay narrow.
Every call to Skyvern — read or write, approved or declined — is recorded with the actor, the input, and the result, and can be linked to the ticket or change record.
Often connected alongside
All integrations
Adobe Marketing Agent MCPAdobe Experience Platform context, sandbox switches, and async task status
AffindaRead parsed documents, collections, and annotations — change them with approval
Agent MailGive the agent its own mailbox for reading threads and sending updates
AI/ML APIInspect assistants, threads, and runs — and change them as approved stepsPut Skyvern to work with Luumen
Connect in minutes. Every action scoped, approved, and audited from day one.