AC model-microbenchmark
Time a vLLM model's forward pass on exact batch shapes (decode, prefill, chunked prefill, prefix hit, spec-decode verify, or mixes) with `vllm bench forward`, single-GPU or DP/TP/EP across nodes. Use for per-step model latency, layout/parallelism comparisons, and checking a model-level optimization. Not for single kernels (use kernel-microbenchmark) or end-to-end serving (use `vllm bench serve`).
Promises to profile a vLLM model's forward pass on exact batch shapes—decode, prefill, chunked prefill, and mixes—using `vllm bench forward` across single or multi-GPU setups with DP/TP/EP. Targets per-step latency, layout and parallelism comparison, and model-level optimization checks.
One file, no scripts. Description spans 1336 tokens and covers intended use cases clearly. No critical findings, quality score 88, grade A. Not for kernel-level profiling (use kernel-microbenchmark) or full serving stacks (use `vllm bench serve`).
Install if you need precise forward-pass latency measurements on specific batch configurations and parallelism strategies.
Time a vLLM model's forward pass on exact batch shapes (decode, prefill, chunked prefill, prefix hit, spec-decode verify, or mixes) with vllm bench forward…
As a process C 57/100 · Has gaps — weak spots: result and completion, inputs and preconditions, failures and branches
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: 1. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
✓ No remarks against the Agent Skills spec
Process rating: all ten parameters 57/100
- 0Result and completion. Does not say what the result is
- 0Inputs and preconditions. Does not say what the process needs to start
- 0Failures and branches. Linear process with no failure handling
- 0Progress reporting. Says nothing while it works
- 50When it triggers. No condition that starts the skill
- 100Tools and files. No external tools needed
- 100Steps. 17 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 1358 tokens
- 100Running it twice. No mutating operations
- low The response is described with custom markup (3 tags): a typed call is more reliable
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
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +4Description says when NOT to use the skill
- +3Description length 399: enough signal without eating the budget
- +4Structure: 4 headings
- +3Step-by-step instructions: 17 items
- +4Has examples (3 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 88.