azure-diagnostics: Add Inspektor Gadget Reference

[mqasimsarfraz]

This PR introduces Inspektor Gadget to allow deeper troubleshooting of AKS clusters. All commands use:

• kubectl debug so no additional installation is required
• MCR images for all the steps
• Content is kept in the references/directory to avoid bloating the initial context

Also, gadget commands are added contextually in networking, node-issues, and pod-failures, with the full reference in references/inspektor-gadget.md.

→ npm run tokens check -- plugin/skills/azure-diagnostics/aks-troubleshooting/references/inspektor-gadget.md

> @github-copilot-for-azure/scripts@1.0.0 tokens
> node --import tsx src/tokens/cli.ts check plugin/skills/azure-diagnostics/aks-troubleshooting/references/inspektor-gadget.md


📊 Token Limit Check
════════════════════════════════════════════════════════════
Files Checked: 1
Files Exceeded: 0

✅ All files within token limits!

cc: @julia-yin

[Copilot]

Pull request overview

This PR adds Inspektor Gadget guidance to the azure-diagnostics AKS troubleshooting skill to enable deeper node/pod-level diagnostics when standard Azure/Kubernetes evidence is inconclusive.

Changes:

• Added a new Inspektor Gadget reference document with command patterns, filters, and a symptom-to-gadget map.
• Introduced a “Deep Diagnostics Flow” reference and added contextual IG command snippets to AKS networking, node-issues, and pod-failures playbooks.
• Updated the main AKS troubleshooting guide to include Inspektor Gadget as the third step in the evidence order.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated 5 comments.

Show a summary per file

File	Description
plugin/skills/azure-diagnostics/aks-troubleshooting/references/inspektor-gadget.md	New IG reference: base command pattern, gadget catalog, and symptom mapping.
plugin/skills/azure-diagnostics/aks-troubleshooting/references/command-flows.md	Adds an IG “Deep Diagnostics Flow” and link to the IG reference.
plugin/skills/azure-diagnostics/aks-troubleshooting/pod-failures.md	Adds IG deep-diagnostics commands for CrashLoopBackOff/OOM-style investigations.
plugin/skills/azure-diagnostics/aks-troubleshooting/node-issues.md	Adds IG guidance for PID pressure / unknown process load.
plugin/skills/azure-diagnostics/aks-troubleshooting/networking.md	Adds IG networking/DNS tracing commands for deeper connectivity troubleshooting.
plugin/skills/azure-diagnostics/aks-troubleshooting/aks-troubleshooting.md	Updates evidence order and guidance to incorporate IG as a third-step diagnostic tool.

[jongio]

A few observations on top of the existing bot review.

Doc-only PR adds Inspektor Gadget commands across four AKS troubleshooting pages plus a new reference. Structure and placement match the existing skill conventions. Three things worth a look below; nothing blocking.

[Copilot]

Pull request overview

Copilot reviewed 7 out of 7 changed files in this pull request and generated 3 comments.

[Copilot]

Pull request overview

Copilot reviewed 7 out of 7 changed files in this pull request and generated no new comments.

[jongio]

One small consistency nit on the new networking deep-diagnostics block. Otherwise the latest pass addresses everything I flagged earlier.

[mqasimsarfraz]

Thanks @jongio for quick reviews! :)

I have addressed the outstanding comment, please feel free to let me know if you have any other thought!

[jongio]

Confirmed - the trace_tcpdrop reference is out of networking.md and the catalog stays the single source of truth. Nothing else from my side. Thanks for the quick turnarounds.

[mqasimsarfraz]

@jongio Thanks for the approval but the changes had conflict because of #1968 so I had to push again! I assume it will require the approval again!

[Copilot]

Pull request overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 5 comments.

[jongio]

Walked through all six files end-to-end. The reference is well-organized - single version pin, clear base command pattern, and the symptom-to-gadget map makes the right gadget easy to find. Contextual IG sections in networking, node-issues, and pod-failures point back to the reference without duplicating it.

No new issues since my last pass. Ship it.

[mqasimsarfraz]

@jongio @tmeschter there were conflicts again :(, I pushed latest changes! The PR is ready to go from myside so please feel free to have a look (hopefully for the last time)!

[jongio]

Re-reviewed after rebase. All feedback from earlier rounds addressed - version pin centralized, trace_tcpdrop removed from inline references, evidence order corrected. Structure is clean: single reference file for the catalog, contextual sections in networking/node-issues/pod-failures don't duplicate it. CI green.