Impact analysis

Overview

Impact analysis evaluates proposed configuration changes against a snapshot to identify affected objects, broken references, and potential risks. The analysis is deterministic and dependency-based, examining how a change to one object ripples through related triggers, automations, macros, and other configuration items.

Analyse a single change

Submit one proposed change for impact analysis. The snapshotId must belong to a connection that your API key has access to.

curl -X POST https://api.configly.app/v1/analysis/impact \
  -H "Authorization: Bearer cly_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "connectionId": "{connectionId}",
    "snapshotId": "{snapshotId}",
    "change": {
      "objectId": 12345,
      "objectType": "trigger",
      "changeType": "modify",
      "virtualData": { "active": false },
      "originalData": { "active": true }
    }
  }'

Request body fields

The response contains the analysis result, including affected objects, broken references, and an overall risk assessment.

Analyse in bulk

Submit up to 100 changes in a single request. Each change is analysed independently against the same snapshot. The endpoint supports partial failure: successful analyses are returned in the analyses array, while any errors appear in the errors array.

curl -X POST https://api.configly.app/v1/analysis/impact/batch \
  -H "Authorization: Bearer cly_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "connectionId": "{connectionId}",
    "snapshotId": "{snapshotId}",
    "changes": [
      {
        "objectId": 12345,
        "objectType": "trigger",
        "changeType": "modify",
        "virtualData": { "active": false }
      },
      {
        "objectId": 67890,
        "objectType": "automation",
        "changeType": "delete"
      }
    ]
  }'

Batch response shape

{
  "data": {
    "analyses": [ ... ],
    "errors": [ ... ]
  },
  "meta": {
    "total": 2,
    "successful": 1,
    "failed": 1
  }
}

The meta object summarises the outcome: total is the number of changes submitted, successful is the count that produced an analysis, and failed is the count that returned an error. When all changes succeed, the errors array is empty.

Error codes

To create virtual changes, see Virtual changes. For AI-generated analysis, see AI analysis.