| name | kernel-analysis-router |
| description | Use when a Linux kernel task is primarily about top-down analysis, path classification, tracepoint work, trace capture, large-log search with rg, or git bisect. This router chooses the correct analysis workflow before detailed code or log analysis starts. |
Kernel Analysis Router
Use this router after kernel-workflow-router when the task is analysis-first.
Choose The Next Skill
- Use
topdown-kernel-diagnosis for top-down recursive analysis where the goal is to classify dominant runtime paths cheaply first, then add targeted stack capture only for hot reasons.
- Use
kernel-tracepoint-pattern when the main task is adding or fixing Linux kernel tracepoints, including TRACE_EVENT, CREATE_TRACE_POINTS, and trace event wiring.
- Use
kernel-trace-analysis-pipeline when the task spans instrumentation, trace collection, trace analysis, and final reporting as one end-to-end pipeline.
- Use
kernel-log-instrumentor when the main gap is insufficient kernel logging and the next step is to add or refine structured logs before broader analysis.
Search Rules
- For large logs, use
rg as the default search tool.
- If the user asks for path attribution or dominant callers, prefer
topdown-kernel-diagnosis before high-overhead tracing.
- If the user already has a trace design and just needs wiring or compile fixes, go directly to
kernel-tracepoint-pattern.
Handoff Rules
- If the task requires booting or controlling a VM before analysis can proceed, switch to
kernel-vm-router first.
- If the task ends in a report or handoff document, apply
handoff-doc-style before final drafting and use experiment-report-writer when the output is leadership-facing or office-style.