Graphsignal
graphsignal/graphsignal
Profile AI inference workloads (vLLM, SGLang, TensorRT-LLM, PyTorch, any GPU application) with the Graphsignal profiler and read the results from its local /signals JSON endpoint.
Analyzes Torch Profiler traces from SGLang, vLLM and TensorRT-LLM servers into kernel attribution, overlap and fusion tables.
Install LLM Torch Profiler Trace Analysis
$ npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install BBuf/AI-Infra-Auto-Driven-SKILLS llm-torch-profiler-analysis --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/llm-torch-profiler-analysis .claude/skills/llm-torch-profiler-analysis && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "llm-torch-profiler-analysis" agent skill from https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/llm-torch-profiler-analysis into .claude/skills/llm-torch-profiler-analysis/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-torch-profiler-analysis", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/llm-torch-profiler-analysisType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install BBuf/AI-Infra-Auto-Driven-SKILLS llm-torch-profiler-analysis --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/llm-torch-profiler-analysis .agents/skills/llm-torch-profiler-analysis && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "llm-torch-profiler-analysis" agent skill from https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/llm-torch-profiler-analysis into .agents/skills/llm-torch-profiler-analysis/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-torch-profiler-analysis", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install BBuf/AI-Infra-Auto-Driven-SKILLS llm-torch-profiler-analysis --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/llm-torch-profiler-analysis .cursor/skills/llm-torch-profiler-analysis && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "llm-torch-profiler-analysis" agent skill from https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/llm-torch-profiler-analysis into .cursor/skills/llm-torch-profiler-analysis/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-torch-profiler-analysis", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS.git --path skills/llm-torch-profiler-analysis--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install BBuf/AI-Infra-Auto-Driven-SKILLS llm-torch-profiler-analysis --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/llm-torch-profiler-analysis .gemini/skills/llm-torch-profiler-analysis && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "llm-torch-profiler-analysis" agent skill from https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/llm-torch-profiler-analysis into .gemini/skills/llm-torch-profiler-analysis/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-torch-profiler-analysis", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install BBuf/AI-Infra-Auto-Driven-SKILLS llm-torch-profiler-analysisInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/llm-torch-profiler-analysis .github/skills/llm-torch-profiler-analysis && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "llm-torch-profiler-analysis" agent skill from https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/llm-torch-profiler-analysis into .github/skills/llm-torch-profiler-analysis/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-torch-profiler-analysis", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install BBuf/AI-Infra-Auto-Driven-SKILLS llm-torch-profiler-analysis --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/llm-torch-profiler-analysis .opencode/skills/llm-torch-profiler-analysis && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "llm-torch-profiler-analysis" agent skill from https://github.com/BBuf/AI-Infra-Auto-Driven-SKILLS/tree/main/skills/llm-torch-profiler-analysis into .opencode/skills/llm-torch-profiler-analysis/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "llm-torch-profiler-analysis", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
llm-torch-profiler-analysis- GitHub stars
- 925
- Token cost
- ~2.8k tokens
- SKILL.md length
- 1,231 words
- Files
- 15 (incl. scripts, references)
- Skills in repo
- 10
- Repo updated
- First seen
- Licence
- None found
At a glance
Analyzes Torch Profiler traces from SGLang, vLLM and TensorRT-LLM servers into kernel attribution, overlap and fusion tables.
- Works in 6 steps: Confirm nonzero GPU events, a complete… → Read all relevant GPU streams. A CPU… → Separate sum of kernel durations, union… → …
- Finding which kernels dominate an LLM decode or prefill trace
- SKILL.md covers Choose the evidence, Read a trace, Capture against a running server and Interpret before optimizing, plus 1 more section
- Runs Python and Shell scripts from its folder; calls python3
What it does
This skill turns a PyTorch profiler trace from an LLM serving framework into three tables: kernel and source attribution, overlap opportunities and fusion patterns. It supports SGLang, vLLM, TensorRT-LLM and TokenSpeed, and recognizes SGLang Omni traces. It is used for questions about model dispatch, CUDA Graph gaps, stream contention and exposed kernel tails.
There are three evidence modes. An existing Chrome JSON or gzip trace with GPU events needs no GPU or framework install, and the Python analyzers use only the standard library. A single live capture uses a shared output directory and a supported HTTP profiler, recording the launch arguments and package revisions. A mapping-plus-formal run takes call-site context from an eager trace and timing from a warmed graph trace, and eager timing is never carried over.
Guidance warns against collapsing distributed runs to rank zero when studying rank skew, and against treating source checks or historical captures as fresh GPU validation. Reference files hold heuristics, fusion and overlap catalogs and a source map, and helper scripts start profiling on SGLang or vLLM hosts.
When your agent uses it
- Finding which kernels dominate an LLM decode or prefill trace
- Looking for CUDA Graph gaps or exposed kernel tails
- Checking a trace for stream contention and overlap opportunities
- Spotting operator fusion candidates in a serving framework's trace
Example prompts
- “Analyze this SGLang decode trace and give me the kernel attribution table.”
- “Capture a profile from the running vLLM server and list the best fusion opportunities.”
- “Why is there a gap between graph replays in this trace? Look for stream contention.”
Requirements
- Python 3, standard library only for the analyzers
- A Torch Profiler Chrome trace, or a running SGLang or vLLM server to capture from
Workflow steps
6 steps, taken from the first numbered list in SKILL.md.
- Confirm nonzero GPU events, a complete warmed forward, actual dispatch and
- Read all relevant GPU streams. A CPU cudaGraphLaunch span is host time;
- Separate sum of kernel durations, union of GPU busy intervals, and end-to-end
- Identify the producer/consumer join and exposed tail. PDL consumers may
- Consult source dispatch before proposing fusion. Existing routed-expert
- Time changes without profiling; run kernel correctness and actual-model
What it can do on your machine
Read from SKILL.md and the folder at commit 6dc9c66. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Ships 9 files in scripts/ (Python and Shell), which the agent can run.
Shell commands in SKILL.md call:
python3
From the folder's file list and the shell code blocks in SKILL.md.
Network
Links to these hosts (documentation or services it may open):
github.com
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
LLM Torch Profiler Trace Analysis loads about 2.8k tokens when it runs, and up to ~451k if it reads all its reference files. Until then it costs about 83 tokens; SKILL.md has 1,231 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.
SKILL.md
Without a licence we can't republish the file, so here is its outline and opening line. It has 1,231 words (~2,828 tokens).
“Produce three tables: kernel/source attribution, overlap opportunities, and fusion patterns. Start from the user's trace or running server. Inspect the actual model, framework revision, phase, device/rank and parallelism before choosing a fast path. Existing traces need no GPU or framework…”
- name
- llm-torch-profiler-analysis
Files
SKILL.md and 14 other files (scripts, references) in skills/llm-torch-profiler-analysis of BBuf/AI-Infra-Auto-Driven-SKILLS.
- SKILL.md
- references/fuse-overlap-catalog.md
- references/heuristics.md
- references/overlap-catalog.md
- references/source-map.md
- references/vllm-torch-compile-fusions.md
- scripts/analyze_llm_torch_profile.py
- scripts/analyze_sglang_torch_profile.py
- scripts/probe_llm_server.py
- scripts/profile_common.py
- scripts/render_triage_markdown_bundle.py
- scripts/run_sglang_torch_profile_host.sh
- scripts/run_vllm_torch_profile_host.sh
- scripts/triage_kernel_helpers.py
- scripts/triage_overlap_helpers.py
Open the folder on GitHubat commit 6dc9c66
Compare with similar skills
LLM Torch Profiler Trace Analysis next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| LLM Torch Profiler Trace Analysis this skillBBuf/AI-Infra-Auto-Driven-SKILLS | 925 | — | ~2.8k | Automated safety check: Pass | None | |
| Graphsignalgraphsignal/graphsignal | 257 | — | ~6.2k | Automated safety check: Pass | Apache-2.0 | |
| Magpie Kernel Evaluatoramd/skills | 406 | — | ~2.3k | Automated safety check: Pass | MIT | |
| LLM Torch Profiler Analysissgl-project/sglang | 37k | 2 repos | ~6.4k | Automated safety check: Pass | Apache-2.0 | |
| Jetson Memory AuditNVIDIA/skills | 3.5k | 1 repos | ~2.3k | Automated safety check: Notes | Apache-2.0 | |
| Jetson PackageNVIDIA/skills | 3.5k | 1 repos | ~1.8k | Automated safety check: Pass | Apache-2.0 |
Similar skills
Magpie Kernel Evaluator
amd/skills
Benchmarks LLM inference and drives GPU kernel optimization with Magpie.
LLM Torch Profiler Analysis
sgl-project/sglang
Unified LLM torch-profiler triage skill for sglang, vllm, TensorRT-LLM, and TokenSpeed.
- Official
Jetson Memory Audit
NVIDIA/skills
Measure Jetson DRAM/NvMap usage and verify before/after memory reclamation with live audit data.
- Official
Jetson Package
NVIDIA/skills
Pick Jetson-compatible containers, vLLM runtime images, and Jetson AI Lab PyPI indexes; maps Orin SM 8.7 vs Thor SM 11.0 and JetPack-specific package choices.
- Official
Jetson Inference Mem Tune
NVIDIA/skills
Pick the serving stack and per-runtime memory flags (vLLM, SGLang, llama.cpp, TensorRT Edge-LLM) for an LLM/VLM workload on any NVIDIA Jetson.
More from BBuf/AI-Infra-Auto-Driven-SKILLS
All 10 skills in this repoSGLang Model Day-0 Support
BBuf/AI-Infra-Auto-Driven-SKILLS
Plans and audits Day-0 SGLang support for a new model release: scope, architecture gaps, PR order, validation gates and sanitized public evidence.
LLM Serving Capacity Planner
BBuf/AI-Infra-Auto-Driven-SKILLS
Reads SGLang or vLLM startup logs to show where GPU memory went and estimates how many concurrent requests fit at common token lengths.
Model Architecture Diagram Finder
BBuf/AI-Infra-Auto-Driven-SKILLS
Looks up public original architecture diagrams for named LLM, vision-language, MoE, diffusion and OCR models and returns the image with its source attribution.
Model Compute Simulator
BBuf/AI-Infra-Auto-Driven-SKILLS
Builds an operator-level compute template for an LLM and estimates FLOPs and MFU for a serving shape, with tensor shapes and parallelism what-if checks.
SGLang Maintainer-Style Review
BBuf/AI-Infra-Auto-Driven-SKILLS
Reviews SGLang changes the way its maintainers do, drawing on a bundled corpus of public PR review threads and a flowchart of how the diff runs.
Torch Profiler Layer Track
BBuf/AI-Infra-Auto-Driven-SKILLS
Adds verified layer guides such as L0 and L1 and compact GPU lanes to an existing Torch Profiler Chrome trace, changing how it looks but not how it ran.
Related
Works with
Categories
Questions about LLM Torch Profiler Trace Analysis
What does LLM Torch Profiler Trace Analysis do?
Analyzes Torch Profiler traces from SGLang, vLLM and TensorRT-LLM servers into kernel attribution, overlap and fusion tables. This skill turns a PyTorch profiler trace from an LLM serving framework into three tables: kernel and source attribution, overlap opportunities and fusion patterns. It supports SGLang, vLLM, TensorRT-LLM and TokenSpeed, and recognizes SGLang Omni traces.
When should I use LLM Torch Profiler Trace Analysis?
LLM Torch Profiler Trace Analysis fits situations like: finding which kernels dominate an LLM decode or prefill trace; looking for CUDA Graph gaps or exposed kernel tails; checking a trace for stream contention and overlap opportunities; spotting operator fusion candidates in a serving framework's trace.
How do I install LLM Torch Profiler Trace Analysis in Claude Code?
Run `npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a claude-code`. Or copy the skill folder (skills/llm-torch-profiler-analysis in BBuf/AI-Infra-Auto-Driven-SKILLS) into .claude/skills/llm-torch-profiler-analysis in your project. Claude Code loads it when a task matches its description.
How do I install LLM Torch Profiler Trace Analysis in Codex?
Run `npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a codex`. Or copy the skill folder (skills/llm-torch-profiler-analysis in BBuf/AI-Infra-Auto-Driven-SKILLS) into .agents/skills/llm-torch-profiler-analysis in your project. Codex loads it when a task matches its description.
Can I use LLM Torch Profiler Trace Analysis in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add BBuf/AI-Infra-Auto-Driven-SKILLS --skill llm-torch-profiler-analysis -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/llm-torch-profiler-analysis, .gemini/skills/llm-torch-profiler-analysis, .github/skills/llm-torch-profiler-analysis and .opencode/skills/llm-torch-profiler-analysis in your project.
What does LLM Torch Profiler Trace Analysis need to run?
Going by SKILL.md and its folder, LLM Torch Profiler Trace Analysis needs Python and a shell for the scripts in its folder and the command-line tools its instructions call (python3). Our summary lists: Python 3, standard library only for the analyzers; A Torch Profiler Chrome trace, or a running SGLang or vLLM server to capture from.
Does LLM Torch Profiler Trace Analysis access the network?
SKILL.md names 1 domain. As links in the text: github.com. This is read from the text; nothing was executed.
Is LLM Torch Profiler Trace Analysis safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
What licence does LLM Torch Profiler Trace Analysis use?
No licence was found for LLM Torch Profiler Trace Analysis or its repository. Without one, default copyright applies: ask the author before reusing or redistributing it.
How many tokens does LLM Torch Profiler Trace Analysis use?
About 2.8k tokens (SKILL.md is roughly 11k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 448k tokens, read only when the agent opens those files.
What are the alternatives to LLM Torch Profiler Trace Analysis?
Skills that share tags, products or a category with LLM Torch Profiler Trace Analysis: Graphsignal (graphsignal/graphsignal, 257 stars), Magpie Kernel Evaluator (amd/skills, 406 stars), LLM Torch Profiler Analysis (sgl-project/sglang, 37k stars) and Jetson Memory Audit (NVIDIA/skills, 3.5k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains LLM Torch Profiler Trace Analysis?
BBuf (a GitHub user) maintains it in BBuf/AI-Infra-Auto-Driven-SKILLS, which has 925 GitHub stars. The repository holds 10 skills in this directory. The repository was last updated on October 5, 2026.
Source: BBuf/AI-Infra-Auto-Driven-SKILLS on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.