Sentry Read-only Observability Skill

Publisher: Sentry plugin in openai/plugins · Editorial profile by AI Tools Ratings · Review method

What task is this skill designed for?

Summarize a defined project and time window, distinguishing event counts from affected users and observation from a proposed root cause. Use issue IDs and event examples to make the brief auditable.

What input and output should you expect?

Input: Organization, project, environment, time range and an incident question. Provide a locally configured read-only token without copying it into a prompt.

Output: A concise issue list, event-level evidence and follow-up investigation steps. A useful brief records filters and missing data so another engineer can reproduce the query.

What does it need to run?

Python and the bundled Sentry API helper, network access and a token with appropriate read scopes. The helper and its request handling should be inspected before use.

Published in the OpenAI Codex plugins repository. Compatibility evidence is the plugin package and its source instructions; this is not a claim of identical behavior in other agent clients.

How should you approach installation?

1. Open the pinned source and review the complete skill folder, including referenced helpers and license, before enabling it.
2. Use the corresponding plugin in a compatible Codex environment; preserve the skill folder structure rather than copying only SKILL.md. Confirm current plugin availability in your client.
3. Start with a disposable project and the trial prompt below. Grant only the stated access, inspect the proposed diff, and require separate approval for remote writes or publishing.

Installation guide: discovery, dependencies and runtime checks

What files, accounts or data can it access?

Requests leave the device for Sentry. Returned stack traces and metadata enter the agent context; redact personal data and avoid broad organization scopes when one project is sufficient.

What does the skill cost?

Sentry plan limits, API access and coding-agent usage apply independently. The source does not bundle a Sentry subscription or guarantee data retention.

What are its limits and alternatives?

Read-only inspection does not resolve incidents, change alerts or prove causation. API pagination and retention constrain coverage; a short recent window can miss a recurring problem.

Use the Sentry interface for interactive triage or a code-review skill when the task is fixing a known defect rather than retrieving telemetry.

What could you ask in a small trial?

Try this with a small sample:

For a sandbox project, summarize unresolved production issues in the past 24 hours. Cite IDs and timestamps, separate evidence from hypotheses, and redact identifiers. Do not change issue status or send messages.

How would you verify the result?

1. The brief states project, environment and time window.
2. Every important finding points to a returned issue or event.
3. No access token or personal payload appears in output.
4. The report distinguishes recurrence from a confirmed regression.

Use a baseline and record acceptance evidence before treating a successful task as a reliable workflow.

Which source and license were checked?

Read the reviewed SKILL.md at its pinned commit · Complete reviewed folder

Repository and current publisher guidance

Checked: 2026-10-30

Apache-2.0 License evidence

Report an error, license concern or request a new review.