tensorrt-llm

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100), when you need 10-100x faster inference than PyTorch, or for serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling.

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

80%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is highly actionable with executable code and good progressive disclosure into real reference files. Its weakest point is workflow clarity: serving and batch-deployment flows lack explicit validation or verification checkpoints.

Suggestions

Add a validation/checklist step to the serving flow (e.g., verify the server is healthy with a curl to /v1/models or /metrics before declaring deployment ready).

For batch inference, include a verification step such as checking output count equals prompt count and inspecting a sample output before reporting results.

Note version-sensitive claims (e.g., 'pip install tensorrt_llm==1.2.0rc3', '24,000 tokens/sec') as benchmark/version snapshots so they don't silently age.

Dimension	Reasoning	Score
Conciseness	Efficient and mostly lean — terse feature bullets and direct code blocks — with only minor instances of restating context Claude already knows (e.g., the opening library description).	4 / 5
Actionability	Provides copy-paste ready, fully executable examples — LLM/SamplingParams inference, trtllm-serve launch, curl client, quantized and multi-GPU configs — covering the common cases.	5 / 5
Workflow Clarity	A clear install → inference → serving sequence exists, but there are no explicit validation checkpoints, and batch inference is mentioned without verification steps for risky/batch operations.	3 / 5
Progressive Disclosure	SKILL.md is a well-organized overview with clearly signaled one-level-deep references to real bundle files (references/optimization.md, multi-gpu.md, serving.md), keeping detail appropriately split out.	5 / 5
	Total	17 / 20 Passed

Description

87%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is strong: it states concrete capabilities, provides explicit 'Use when...' trigger guidance, and is clearly scoped to NVIDIA production LLM serving. Minor keyword synonym coverage is the only gap.

Dimension	Reasoning	Score
Specificity	Names the domain and several concrete capabilities — 'production deployment,' 'serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling' — with only minor gaps in coverage.	4 / 5
Completeness	Explicitly answers both 'what' (Optimizes LLM inference...) and 'when' ('Use for production deployment... when you need 10-100x faster inference... or for serving models...') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Includes natural user-facing terms like 'production deployment on NVIDIA GPUs (A100/H100),' 'quantization (FP8/INT4),' 'in-flight batching,' and 'multi-GPU scaling,' though a few synonyms are missing.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear NVIDIA TensorRT production-serving niche with distinct hardware/feature triggers, minimizing overlap with adjacent tools like vLLM or llama.cpp.	5 / 5
	Total	18 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: OpenLAIR/dr-claw
Path: skills/inference-serving/tensorrt-llm/SKILL.md
Commit: 4b35756

Reviewed: 1 day ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.