← Explore all skills

CLOUD OPERATIONS / JetBrains

Production Incident Responder

Runs production incident response as mitigate-first: state impact, pick the fastest reversible mitigation, preserve evidence while the symptom still exists, then diagnose — with heuristics for retry storms, cache stampedes and config-only regressions.

“Guide production incident response for Kotlin plus Spring services from first alert through mitigation, diagnosis, and follow-up.”

01 / THE REASONING

Why this made the selection.

  • Orders the response deliberately — reversible mitigation before root-cause work — and names the anti-patterns: reflexive restarts that destroy evidence and amplify connection storms, scaling the app tier when the database is the bottleneck.
  • The heuristics name specific failure patterns: a healthy dependency at low volume failing under your own fleet's retry storm, config-only incidents masquerading as code regressions, rollback being unsafe after data shape has changed.
  • The output contract separates confirmed facts from working hypotheses and requires impact, mitigation, evidence and next diagnostic step.

02 / THE REVIEW RECORD

What we actually inspected.

Source review has boundaries.
A clear record is more useful than a “safe” badge.

Material inspected

  • production-incident-responder/SKILL.md (JetBrains mirror, complete, 96 lines)
  • production-incident-responder/agents/openai.yaml (identical to upstream)
  • upstream yalishevant/kotlin-backend-agent-skills production-incident-responder SKILL.md
  • upstream yalishevant/kotlin-backend-agent-skills LICENSE
  • JetBrains/skills README mirror policy
  • https://github.com/Kotlin/kotlin-backend-agent-skills (answers HTTP 301 to yalishevant/kotlin-backend-agent-skills)
  • mirror SKILL.md diffed against the upstream SKILL.md at fd0ed521d22efd28221d532c9abab97af0b9bf0e

Our findings

  • First principles make mitigation reversible-action-first and evidence preservation an explicit step before diagnosis begins.
  • Guardrails forbid code changes as the first incident action when a reversible mitigation exists, and root-cause certainty without evidence.
  • A diff against the upstream SKILL.md at fd0ed521d22efd28221d532c9abab97af0b9bf0e shows the mirror adds four frontmatter lines (metadata, short-description, author: Kotlin, source) and nothing else; the upstream file has no author field. agents/openai.yaml is identical in both.
  • Upstream health: https://github.com/Kotlin/kotlin-backend-agent-skills, which the mirror’s metadata.source and README name, now answers HTTP 301 to yalishevant/kotlin-backend-agent-skills, a personal account (https://api.github.com/repos/Kotlin/kotlin-backend-agent-skills returns full_name yalishevant/kotlin-backend-agent-skills, owner type User). fd0ed521d22efd28221d532c9abab97af0b9bf0e is still that repository’s head, and its LICENSE (MIT, copyright 2026 Anton Yalyshev) is unchanged (checked 2026-10-13).

Not established by this review

  • No incident was responded to and no mitigation was executed; the workflow was read, not run.
  • Heuristic claims about retry storms and cache stampedes were not reproduced.

The review applies to the material and revision named here. A newer upstream release can change its behavior.

03 / PUT IT TO WORK

Add a skill to your workflow.

Upstream setup instructions ↗

DOCUMENTED COMMAND

npx skills add JetBrains/skills/production-incident-responder --skill production-incident-responder

Run in your terminal to choose your AI tools, project or global scope, and installation method. This short command follows the current upstream version; the review covers the revision shown above.

Optional: install the reviewed revision

Keeps the original source and installer version where recorded, while leaving your AI tools and installation scope for you to choose.

npx skills add https://github.com/JetBrains/skills/tree/e0f258b5cfed145015cb3e48da9a97947f7c4ed7/production-incident-responder --skill production-incident-responder
  1. Install the complete skill directory; the agents profile travels with the SKILL.md.
  2. Bring the alert state, dashboards and recent-change timeline; expect a mitigation-first response with an explicit evidence and hypothesis log.

Before you start

  • Node.js/npm for the installer
  • An agent that supports the Agent Skills format
  • Access to alerts, dashboards, traces and the project's rollback or feature-flag mechanisms

04 / FOLLOW THE EVIDENCE

The source trail.

Our notes are separate from the original resource.
Check upstream before adopting a new version.

  1. https://github.com/JetBrains/skills/blob/e0f258b5cfed145015cb3e48da9a97947f7c4ed7/production-incident-responder/SKILL.md

    Supports: summary, upstreamDescription, whySelected, bestFor, limitations, review, access

  2. https://github.com/yalishevant/kotlin-backend-agent-skills/blob/fd0ed521d22efd28221d532c9abab97af0b9bf0e/.agents/skills/production-incident-responder/SKILL.md

    Supports: review, limitations

  3. Upstream licence (MIT) ↗Checked

    https://github.com/yalishevant/kotlin-backend-agent-skills/blob/fd0ed521d22efd28221d532c9abab97af0b9bf0e/LICENSE

    Supports: license

  4. https://github.com/Kotlin/kotlin-backend-agent-skills

    Supports: review, limitations

  5. https://github.com/JetBrains/skills/blob/e0f258b5cfed145015cb3e48da9a97947f7c4ed7/README.md

    Supports: review, install, limitations

  6. https://github.com/vercel-labs/skills/blob/7407f3893ad4dceab546ac002c3ef806e4000c73/README.md

    Supports: install

Evidence & Ask