Skip to content

fix: harden OHE troubleshooting skill for customer use - #457

Draft
rajshah4 wants to merge 1 commit into
OpenHands:skill/openhands-enterprise-troubleshootingfrom
rajshah4:harden-ohe-troubleshooting-skill
Draft

fix: harden OHE troubleshooting skill for customer use#457
rajshah4 wants to merge 1 commit into
OpenHands:skill/openhands-enterprise-troubleshootingfrom
rajshah4:harden-ohe-troubleshooting-skill

Conversation

@rajshah4

@rajshah4 rajshah4 commented Aug 5, 2026

Copy link
Copy Markdown
Member

HUMAN: This is a skill i have reviewed / run to help users troubleshoot the replicated install. There is a balance between keeping an eye on the system and not go too far into the Kubernetes.

Summary

Updates the skill introduced by #456 with a customer-safe, version-aware troubleshooting workflow.

  • makes the initial diagnostic pass read-only
  • prohibits credential decoding and complete Secret output
  • requires explicit approval, backup checks, impact disclosure, and a recovery path before mutations
  • discovers OHE, Embedded Cluster, Kubernetes, namespace, and workload topology before targeting resources
  • corrects current runtime-api, runtime-*, Keycloak, kotsadm, and Embedded Cluster guidance
  • uses the supported application-installer support-bundle workflow
  • adds support-bundle privacy, comparison, and escalation guidance
  • adds the plugin manifest, Claude/Codex vendor symlinks, generated catalog entry, and marketplace test count missing from the original draft

This PR intentionally targets skill/openhands-enterprise-troubleshooting, the head branch for #456. Merging it updates that draft without creating a competing implementation against main.

Evidence

Validated against a live installation running OHE 0.36.1 and Embedded Cluster v2.19.2+k8s-1.36 using read-only checks for installer version/help, KOTS release discovery, namespaces, workloads, TLS Secret metadata, and runtime discovery.

python3 skills/skill-creator/scripts/quick_validate.py skills/openhands-enterprise-troubleshooting
Skill is valid!

uv run --group test pytest -q
540 passed, 6 skipped

All 26 Bash examples also passed syntax and executable safety scans, and the changed skill passed a credential-pattern scan.

Scope

Customer-specific domains, IDs, credentials, paths, and one-off recovery patches were excluded. Generic patterns learned from operational troubleshooting and support-bundle triage were retained.

This pull request was created by OpenHands on behalf of Rajiv Shah.

Co-authored-by: openhands <openhands@all-hands.dev>
@rajshah4 rajshah4 changed the title Harden OHE troubleshooting skill for customer use fix: harden OHE troubleshooting skill for customer use Aug 5, 2026
@github-actions github-actions Bot added the type: fix A bug fix label Aug 5, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

type: fix A bug fix

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant