pilot-rollback-decision

Evaluate ConfigHub history and receipts to select the safest Kubernetes rollback revision.

14|Updated Jan 17, 2026
One-click install
npx skills add https://github.com/confighub/cub-scout --skill pilot-rollback-decision
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pilot-rollback-decision
Source: https://github.com/confighub/cub-scout/tree/main/skills/pilot-rollback-decision
Command: npx skills add https://github.com/confighub/cub-scout --skill pilot-rollback-decision

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you decide whether to roll back a broken Kubernetes release and which prior revision is safest to target.

Core Features & Use Cases

  • Rollback target safety verdicts: Evaluates candidate prior revisions with receipt verify --at-commit, producing PASS/WATCH/BLOCK/INCONCLUSIVE guidance.
  • Evidence-backed decisioning: Uses ConfigHub history (ChangeSet timeline) plus impact (blast-radius preview) and receipt data to explain why a rollback target is chosen.
  • Decision chain for auditability: Optionally chains per-candidate receipts into a chosen-target receipt via --input-attestation, preserving the full evaluation rationale.

Quick Start

Ask the AI to produce a rollback verdict for Deployment/payments-api in prod by evaluating candidate revisions from history and selecting the most-recent PASS target.

Frequently Asked Questions about pilot-rollback-decision

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I decide which Kubernetes revision is safest to roll back to after a broken release?

To determine the safest Kubernetes rollback target, evaluate candidate prior revisions using receipt verification and blast-radius impact analysis. This produces a structured safety verdict—PASS, WATCH, BLOCK, or INCONCLUSIVE—for each candidate, identifying the most recent revision that passed verification.

What is a rollback safety verdict and how does it work with GitOps?

A rollback safety verdict is a structured assessment that determines if a prior Kubernetes revision is safe to restore. It works by verifying revision-pinned receipts at each candidate commit and analyzing ConfigHub history to produce auditable PASS or BLOCK guidance for GitOps operations.

How do I evaluate blast-radius impact before rolling back a Kubernetes deployment?

Evaluating blast-radius impact before a Kubernetes rollback involves previewing the potential effect of reverting to a candidate revision. By analyzing ConfigHub history and impact data alongside per-candidate receipts, you receive evidence-backed guidance on whether a target is safe or should be blocked.

Can I audit the decision chain when choosing a rollback target in Kubernetes?

Yes, you can audit the rollback decision chain by chaining per-candidate evaluation receipts into a final chosen-target receipt. Using input attestation preserves the full safety verdict rationale, ensuring the evidence behind a Kubernetes rollback target is completely traceable for post-incident reviews.

Do I need ConfigHub connected-mode capabilities to verify Kubernetes rollback candidates?

Yes, generating rollback safety verdicts requires ConfigHub connected-mode capabilities for history and impact analysis. These features enable the ChangeSet timeline review and blast-radius preview necessary to evaluate candidate revisions and produce a verified rollback target receipt.

When should I not use an automated rollback target decision for my Kubernetes deployment?

You should avoid automated rollback target decisions when verification results are INCONCLUSIVE or BLOCK. If candidate revisions fail receipt verification or the blast-radius impact is unacceptable, the system prevents a rollback, requiring manual investigation before reverting the broken Kubernetes release.