GET /v1/analyses/{id}/similar

Find past analyses similar to this one — root cause, services, and severity matching

Returns past analyses in your tenant that resemble the given one, ranked by similarity score. Useful for pattern-matching ("have we seen this before?") and for finding the previous fix for a recurring issue.

Plan gate. This endpoint is available on Pro and above. Starter accounts receive 403 FEATURE_NOT_AVAILABLE.

Authentication

API key with analysis:read permission.

Endpoint

GET /v1/analyses/{id}/similar

Path parameters

Param	Type	Notes
`id`	UUID	The analysis id

Response

200 OK

{
  "success": true,
  "data": [
    {
      "id": "ana_a1b2...",
      "summary": "Connection pool exhaustion in checkout-api...",
      "severity": "high",
      "createdAt": "2026-04-12T09:14:00Z",
      "similarity": 0.87,
      "feedbackAccuracy": "correct"
    }
  ]
}

Field	Type	Notes
`similarity`	`number` (0–1)	Match score; higher = more similar
`feedbackAccuracy`	string \| null	The feedback rating on the matched analysis (if any)

The list is sorted descending by similarity.

Error responses

Status	`error.code`	When
400	`VALIDATION_ERROR`	`id` is not a UUID
401	`UNAUTHORIZED`	Missing or invalid API key
403	`FORBIDDEN`	Key lacks `analysis:read`
403	`FEATURE_NOT_AVAILABLE`	Plan tier doesn't include similar incidents (Starter)
404	`NOT_FOUND`	No analysis with that id

Example

curl https://api.sigsentry.com/v1/analyses/ana_x1y2/similar \
  -H "Authorization: Bearer ss_secret_..."

How matching works

Matches are ranked by a similarity score from 0 to 1. Raw log content is not compared across analyses. See Similar incidents for the conceptual walkthrough.

Privacy

Matching is scoped to your tenant. Analyses across tenants are never compared and a match never reveals data from another tenant.

GET /v1/analyses/{id}/similar

On this page