Connect a client
Read deployment logs
Read a build or deployment log in chunks of up to 32,768 characters. Pass nextCursor back as cursor to continue.
Arguments
| Name | Type | Required | Limits |
|---|---|---|---|
cursor | string | No | See schema |
limit | integer | No | Minimum 1; Maximum 32768 |
projectId | string | Yes | At least 1 character(s) |
deploymentId | string | Yes | At least 1 character(s) |
Full input schema
{
"type": "object",
"properties": {
"cursor": {
"type": "string"
},
"limit": {
"type": "integer",
"minimum": 1,
"maximum": 32768
},
"projectId": {
"type": "string",
"minLength": 1
},
"deploymentId": {
"type": "string",
"minLength": 1
}
},
"required": [
"projectId",
"deploymentId"
],
"additionalProperties": false
}Example call
The params of a tools/call request. Replace the placeholder values with your own IDs.
{
"name": "getDeploymentLogs",
"arguments": {
"projectId": "REPLACE_WITH_PROJECT_ID",
"deploymentId": "REPLACE_WITH_DEPLOYMENT_ID"
}
} Result
The same data as the API operation, wrapped as {"untrusted_data": true, "source": "getDeploymentLogs", "note": "...", "data": ...} so the model treats project content as data, not instructions.