AC perf-host-analysis
Analyze host/CPU overhead in TensorRT-LLM inference from nsys traces. Detect whether host overhead is the bottleneck using GPU idle ratio, host prep exposed ratio, and per-phase evidence. For regressions, isolate forward steps via allreduce/NVTX patterns, compare host operation breakdowns across versions, and identify scheduling or request-management overhead. Supports optional inter-kernel gap, eager-vs-graph, pattern mapping, and multi-rank straggler drill-down. Use standalone or within perf-analysis. Triggers: host overhead, inter-step gap, scheduling overhead, forward step isolation, nsys iteration analysis, NVTX breakdown, request management overhead, GPU idle, host bottleneck, host prep exposed, inter-kernel gap, bubble analysis, graph coverage, eager kernel, rank imbalance, straggler detection.
This skill analyzes nsys profiles from TensorRT-LLM to spot host bottlenecks via GPU idle ratio and host prep exposed ratio. Promises to isolate forward steps, compare versions, and catch scheduling overhead and stragglers across ranks.
Contains 12 files, 2 scripts, 3972 code tokens. Quality score 92, safety 100, no critical findings. Built for nsys traces and NVTX markers—without them it won't help. Runs on all major platforms including Claude and Cursor. Process score of 64 signals the logic is there but complex and demands clean input data.
Install if you profile LLM inference and hunt regressions. Without nsys and NVTX markup in your code, it's dead weight.
Analyze host/CPU overhead in TensorRT-LLM inference from nsys traces.
As a process C 64/100 · Has gaps — weak spots: result and completion, progress reporting
How to improve
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 12. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- note
frontmatter-keyunknown frontmatter key "dependencies"
Process rating: all ten parameters 64/100
- 0Result and completion. Does not say what the result is
- 0Progress reporting. Says nothing while it works
- 50When it triggers. No condition that starts the skill
- 60Tools and files. Uses tools (web, python) that frontmatter does not declare
- 60Failures and branches. 2 branches
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Steps. 38 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 3970 tokens
- 100Running it twice. No mutating operations
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +3Description length 812: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- +2Single-language instructions
- +4Description says when NOT to use the skill
- +4Structure: 40 headings
- +3Step-by-step instructions: 38 items
- +4Has examples (10 code blocks)
- +4Reference files are cited in the instructions (9 of 9)
- +3All 2 scripts are documented
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 92.