Notes on the Sentry Web API

Posted on In ,

These notes focus on Sentry’s web API, not server installation.

Official docs: https://docs.sentry.io/api/

Community: https://forum.sentry.io/

As of May 2017, the v0 API is still in draft—endpoints may change.

Authentication

All requests must include an auth header:

1
2
headers = {'Authorization': 'Bearer TOKEN'}
response = requests.get(api_url, headers=headers)

Missing credentials yield {"detail": "Authentication credentials were not provided."}; bad tokens return {"detail": "Invalid token"}. Generate tokens under Settings → API → Auth Tokens.

Example API

Sentry offers several categories. Take Events → List a Project’s Events as an example:

1
GET /api/0/projects/{organization_slug}/{project_slug}/events/
  • Method: GET
  • URL: <your-sentry>/api/0/projects/<org_slug>/<project_slug>/events/

organization_slug is the org slug, project_slug the project slug.

Not every endpoint uses GET; deletions use DELETE, updates use PUT, and some require additional headers such as Content-Type: application/json.

Handling responses

The endpoint returns at most 100 events per page. Parse with json.loads(response.text) to get a list of dicts.

Pagination info lives in the Link header:

1
Link: <...cursor=...>; rel="previous"; results="false", <...cursor=...>; rel="next"; results="true"

Fetch the Link header (response.headers.get('Link')), extract the rel="next" URL, and repeat until there is no next link or the page is empty.

Notes

  • The web UI hits internal endpoints; payloads may differ from the public API (e.g., event IDs are longer).
  • The documentation is incomplete—expect to experiment.