Agent skill

Torchscan

Inspect and compare PyTorch models with TorchScan reports, operator FLOPs, and peak-memory workloads.

Apache-2.0Auto-check passedAI & LLM Engineering

Install Torchscan

skills CLI

$ npx skills add frgfm/torch-scan --skill torchscan -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI

$ gh skill install frgfm/torch-scan torchscan --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy

$ git clone --depth 1 https://github.com/frgfm/torch-scan.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/torchscan .claude/skills/torchscan && rm -rf skills-src

Use ~/.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/

Paste into Claude Code

Install the "torchscan" agent skill from https://github.com/frgfm/torch-scan/tree/main/.agents/skills/torchscan into .claude/skills/torchscan/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "torchscan", 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

$skill-installer install https://github.com/frgfm/torch-scan/tree/main/.agents/skills/torchscan

Type 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.

skills CLI

$ npx skills add frgfm/torch-scan --skill torchscan -a codex

Project install goes to .agents/skills/; add -g for ~/.codex/skills/.

GitHub CLI

$ gh skill install frgfm/torch-scan torchscan --agent codex

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/frgfm/torch-scan.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.agents/skills/torchscan .agents/skills/torchscan && rm -rf skills-src

Use ~/.agents/skills/ instead of .agents/skills for a personal install.

Codex skills documentation · loads skills from .agents/skills/

Paste into Codex

Install the "torchscan" agent skill from https://github.com/frgfm/torch-scan/tree/main/.agents/skills/torchscan into .agents/skills/torchscan/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "torchscan", 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.

skills CLI

$ npx skills add frgfm/torch-scan --skill torchscan -a cursor

Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.

GitHub CLI

$ gh skill install frgfm/torch-scan torchscan --agent cursor

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/frgfm/torch-scan.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.agents/skills/torchscan .cursor/skills/torchscan && rm -rf skills-src

Use ~/.cursor/skills/ instead of .cursor/skills for a personal install.

Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/

Paste into Cursor

Install the "torchscan" agent skill from https://github.com/frgfm/torch-scan/tree/main/.agents/skills/torchscan into .cursor/skills/torchscan/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "torchscan", 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

$ gemini skills install https://github.com/frgfm/torch-scan.git --path .agents/skills/torchscan

--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.

skills CLI

$ npx skills add frgfm/torch-scan --skill torchscan -a gemini-cli

Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.

GitHub CLI

$ gh skill install frgfm/torch-scan torchscan --agent gemini-cli

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/frgfm/torch-scan.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.agents/skills/torchscan .gemini/skills/torchscan && rm -rf skills-src

Use ~/.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/

Paste into Gemini CLI

Install the "torchscan" agent skill from https://github.com/frgfm/torch-scan/tree/main/.agents/skills/torchscan into .gemini/skills/torchscan/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "torchscan", 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.

GitHub CLI

$ gh skill install frgfm/torch-scan torchscan

Installs 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).

skills CLI

$ npx skills add frgfm/torch-scan --skill torchscan -a github-copilot

Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.

Manual copy

$ git clone --depth 1 https://github.com/frgfm/torch-scan.git skills-src && mkdir -p .github/skills && cp -r skills-src/.agents/skills/torchscan .github/skills/torchscan && rm -rf skills-src

Use ~/.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/

Paste into GitHub Copilot

Install the "torchscan" agent skill from https://github.com/frgfm/torch-scan/tree/main/.agents/skills/torchscan into .github/skills/torchscan/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "torchscan", 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.

skills CLI

$ npx skills add frgfm/torch-scan --skill torchscan -a opencode

OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.

GitHub CLI

$ gh skill install frgfm/torch-scan torchscan --agent opencode

Project scope by default (.agents/skills/); add --scope user for a personal install.

Manual copy

$ git clone --depth 1 https://github.com/frgfm/torch-scan.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.agents/skills/torchscan .opencode/skills/torchscan && rm -rf skills-src

Use ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.

OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/

Paste into OpenCode

Install the "torchscan" agent skill from https://github.com/frgfm/torch-scan/tree/main/.agents/skills/torchscan into .opencode/skills/torchscan/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "torchscan", 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: torchscan
GitHub stars: 223
Token cost: ~1k tokens
SKILL.md length: 425 words
Files: 1
Skills in repo: 1
Repo updated: Oct 8, 2026
First seen: Oct 7, 2026
Licence: Apache-2.0

At a glance

Inspect and compare PyTorch models with TorchScan reports, operator FLOPs, and peak-memory workloads.

Works in 6 steps: Reuse the project's model and… → Prefer args and kwargs for real calls;… → Use strict=True when incomplete module… → …
An agent must analyze model structure
SKILL.md covers Workflow and Truth rules
Reaches frgfm.github.io

What it does

Torchscan is an agent skill from frgfm/torch-scan. Inspect and compare PyTorch models with TorchScan reports, operator FLOPs, and peak-memory workloads. Use when an agent must analyze model structure, parameters, compute, memory, regressions, unsupported operations, or an owner-provided model budget without inventing completeness.

Its SKILL.md is about 1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts. Compatibility notes: Requires Python 3.11+, PyTorch 2.1+, and the torchscan package. Accelerator claims require matching real hardware.

It sits in AI & LLM Engineering, covering Deep learning. It works with PyTorch and Python. The repository describes itself as: Seamless analysis of your PyTorch models (RAM usage, FLOPs, MACs, receptive field, etc.). The licence is Apache-2.0.

When your agent uses it

An agent must analyze model structure
Unsupported operations
An owner-provided model budget without inventing completeness

Example prompts

“/torchscan”

Requirements

Python 3
Compatibility (from SKILL.md): Requires Python 3.11+, PyTorch 2.1+, and the torchscan package. Accelerator claims require matching real hardware.

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

Reuse the project's model and representative inputs. Do not download weights without permission.
Prefer args and kwargs for real calls; use input_shape only for simple synthetic tensors.
Use strict=True when incomplete module metrics must stop automation.
Serialize the report directly. Never parse the summary table.
Check every metric's status and preserve diagnostics.
Ask the owner for thresholds. TorchScan measures; it does not decide whether a model fits.

What it can do on your machine

Read from SKILL.md and the folder at commit 73fbc11. 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
No scripts in the folder and no shell commands in SKILL.md.
From the folder's file list and the shell code blocks in SKILL.md.
Network
Hosts in commands or code, which the agent is likely to contact:
- frgfm.github.io
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.
Compatibility
Requires Python 3.11+, PyTorch 2.1+, and the torchscan package. Accelerator claims require matching real hardware.
From compatibility in the SKILL.md frontmatter.

Context cost

Torchscan loads about 1k tokens when it runs. Until then it costs about 73 tokens; SKILL.md has 425 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it: ~73
When it runs · the whole SKILL.md, loaded when a task matches: ~1k

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); files beside SKILL.md are not scanned.

SKILL.md

The full file from frgfm/torch-scan at commit 73fbc11, republished under its Apache-2.0 licence (© frgfm). 425 words, ~1,021 tokens.

Download SKILL.mdSave it as .claude/skills/torchscan/SKILL.md (or your agent's skills folder).

name: torchscan
description: Inspect and compare PyTorch models with TorchScan reports, operator FLOPs, and peak-memory workloads. Use when an agent must analyze model structure, parameters, compute, memory, regressions, unsupported operations, or an owner-provided model budget without inventing completeness.
compatibility: Requires Python 3.11+, PyTorch 2.1+, and the torchscan package. Accelerator claims require matching real hardware.
license: Apache-2.0
metadata.author: frgfm
metadata.version: 0.2

TorchScan

Use the smallest API that answers the request:

crawl_module(...): JSON-serializable module report.
summary(...): printed table plus the same report.
mode="structure" on either API: hierarchy, shapes, calls, parameters, and buffers with less overhead.
measure_flops(workload): operator FLOPs for one zero-argument workload call.
measure_peak_memory(workload, device=...): backend-specific PyTorch peak memory.
measure_peak_rss(command): Linux/macOS child-process lifetime RSS, including loading and imports.
profile_workload(workload, device=...): one instrumented operator diagnostic pass; not clean latency.
measure_latency(workload, device=..., inputs=...): first-call time, warmed block-average timing, and explicit work-unit throughput. Unreleased; install main. The callable is invoked repeatedly and owns its state.
compare_reports(before, after): pure same-schema comparison.
compare_benchmarks(before, after, check=...): compatible workload comparison with an owner-supplied output check.
render_report(report): offline model HTML/SVG or benchmark/comparison HTML, without remeasurement.

Workflow

Reuse the project's model and representative inputs. Do not download weights without permission.
Prefer args and kwargs for real calls; use input_shape only for simple synthetic tensors.
Use strict=True when incomplete module metrics must stop automation.
Serialize the report directly. Never parse the summary table.
Check every metric's status and preserve diagnostics.
Ask the owner for thresholds. TorchScan measures; it does not decide whether a model fits.

Truth rules

complete: use value with its method, unit, scope, and context.
partial: known_value is only a lower bound; do not extrapolate.
unavailable: report that no measurement was produced.
Zero is valid only with status == "complete".
Structure mode's compute totals have method not_requested; strict checks cover requested metrics only.
Keep module FLOPs and operator FLOPs separate.
Peak PyTorch memory is not process RSS or total device memory.
Mocked or skipped CUDA/MPS checks are not hardware evidence.
Timing inputs are caller-supplied metadata. Block-average latency is not request p95, and first-call time is not model loading or fresh-process startup. Use compare_benchmarks for timing and compare_reports for model estimates.
A passed output check does not establish task accuracy. Failed checks withhold benchmark deltas; IQR labels are descriptive, not statistical significance. Preserve methods, memory scopes, hardware, and raw timing evidence.

Show full SKILL.md (100 more words)Show less

For an uncounted operator, preserve the partial result. Supply custom_mapping to crawl_module, summary, or measure_flops only when the owner can justify that operator's counting convention. For custom module estimates, use per-analysis custom_modules={ModuleType: ModuleHandler(callback)}. Callbacks receive a complete ModuleCall; declare inclusive subtree ownership per metric to avoid double-counting children. Keep the module and operator views separate. Do not create a global registry, baseline store, wrapper service, or automatic budget policy.

In a repository checkout, read ../../../docs/docs/agent-quickstart.md for the full workflow and ../../../docs/docs/report-schema.md for the report contract, and ../../../docs/docs/extensions.md for copyable extension examples. Outside a checkout, use the published documentation at https://frgfm.github.io/torch-scan/.

Files

Just SKILL.md in .agents/skills/torchscan of frgfm/torch-scan.

Open the folder on GitHubat commit 73fbc11

Compare with similar skills

Torchscan 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.

Torchscan compared with similar skills
Skill	Stars	Used in	Tokens	Auto-check	Licence	Repo updated
Torchscan this skillfrgfm/torch-scan	223	—	~1k	Automated safety check: Pass	Apache-2.0	today
Benchmark Pyreflyfacebook/pyrefly	7.1k	—	~1.8k	Automated safety check: Pass	MIT	today
Document Public APIspytorch/pytorch	104k	—	~4.2k	Automated safety check: Pass	Custom licence	today
ExecuTorch Cortex-M Backendpytorch/executorch	5.1k	—	~872	Automated safety check: Pass	Custom licence	today
PyTorch Lightning TrainingOrchestra-Research/AI-Research-SKILLs	13k	7 repos	~2.3k	Automated safety check: Pass	MIT	3 mo ago
Ako4allTongmingLAIC/AKO4ALL	369	—	~4k	Automated safety check: Pass	MIT	23 days ago

Similar skills

Benchmark Pyrefly
facebook/pyrefly
Official
Run Pyrefly benchmarks locally via Buck or Cargo, including PyTorch real-world LSP benchmarks.
7.1k GitHub stars~1.8k tokensUpdated today
AI & LLM EngineeringAuto-check passed
Document Public APIs
pytorch/pytorch
Document undocumented public APIs in PyTorch by removing functions from coverageignorefunctions and coverageignoreclasses in docs/source/conf.py, running Sphinx coverage, and adding the appropriate…
104k GitHub stars~4.2k tokensUpdated today
AI & LLM EngineeringAuto-check passed
ExecuTorch Cortex-M Backend
pytorch/executorch
Developer guide for the Cortex-M (CMSIS-NN) backend in ExecuTorch: quantization pipeline, pass manager, tests and adding new ops.
5.1k GitHub stars~872 tokensUpdated today
AI & LLM EngineeringAuto-check passed
PyTorch Lightning Training
Orchestra-Research/AI-Research-SKILLs
Shows how to organize PyTorch training with Lightning's LightningModule and Trainer, covering validation, DDP, callbacks and learning-rate scheduling.
13k GitHub starsUsed in 7 repos~2.3k tokens
AI & LLM EngineeringAuto-check passed
Ako4all
TongmingLAIC/AKO4ALL
Drive an agentic loop that iteratively optimizes a GPU kernel for maximum speedup.
369 GitHub stars~4k tokensUpdated 23 days ago
AI & LLM EngineeringAuto-check passed
Homepage Generator
wanshuiyin/ARIS-in-AI-Offer
Generate a fact-checked academic personal homepage from a CV, optionally augmented by an existing manual homepage and an assets directory.
580 GitHub stars~4.8k tokensUpdated 2 days ago
AI & LLM EngineeringAuto-check: notes

Works with

Questions about Torchscan

What does Torchscan do?

Inspect and compare PyTorch models with TorchScan reports, operator FLOPs, and peak-memory workloads. Torchscan is an agent skill from frgfm/torch-scan. Inspect and compare PyTorch models with TorchScan reports, operator FLOPs, and peak-memory workloads.

When should I use Torchscan?

Torchscan fits situations like: an agent must analyze model structure; unsupported operations; an owner-provided model budget without inventing completeness.

How do I install Torchscan in Claude Code?

Run `npx skills add frgfm/torch-scan --skill torchscan -a claude-code`. Or copy the skill folder (.agents/skills/torchscan in frgfm/torch-scan) into .claude/skills/torchscan in your project. Claude Code loads it when a task matches its description.

How do I install Torchscan in Codex?

Run `npx skills add frgfm/torch-scan --skill torchscan -a codex`. Or copy the skill folder (.agents/skills/torchscan in frgfm/torch-scan) into .agents/skills/torchscan in your project. Codex loads it when a task matches its description.

Can I use Torchscan 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 frgfm/torch-scan --skill torchscan -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/torchscan, .gemini/skills/torchscan, .github/skills/torchscan and .opencode/skills/torchscan in your project.

What does Torchscan need to run?

SKILL.md names no scripts, command-line tools or credentials: Torchscan is instructions for the agent only. Our summary lists: Python 3. Compatibility (from SKILL.md): Requires Python 3.11+, PyTorch 2.1+, and the torchscan package. Accelerator claims require matching real hardware..

Does Torchscan access the network?

SKILL.md names 1 domain. In commands or code: frgfm.github.io; the agent is likely to contact it when it follows the instructions. This is read from the text; nothing was executed.

Is Torchscan 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. Review the folder before installing.

What licence does Torchscan use?

Torchscan is published under the Apache-2.0 licence (declared in SKILL.md). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Torchscan use?

About 1k tokens (SKILL.md is roughly 4.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Torchscan?

Skills that share tags, products or a category with Torchscan: Benchmark Pyrefly (facebook/pyrefly, 7.1k stars), Document Public APIs (pytorch/pytorch, 104k stars), ExecuTorch Cortex-M Backend (pytorch/executorch, 5.1k stars) and PyTorch Lightning Training (Orchestra-Research/AI-Research-SKILLs, 13k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Torchscan?

frgfm (a GitHub user) maintains it in frgfm/torch-scan, which has 223 GitHub stars. The repository was last updated on October 8, 2026.

Source: frgfm/torch-scan on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.

Install Torchscan

Facts

At a glance

What it does

When your agent uses it

Example prompts

Requirements

Workflow steps

What it can do on your machine

Tool permissions

Runs code

Network

Credentials

Compatibility

Context cost

Safety

SKILL.md

TorchScan

Workflow

Truth rules

Files

Compare with similar skills

Similar skills

Benchmark Pyrefly

Document Public APIs

ExecuTorch Cortex-M Backend

PyTorch Lightning Training

Ako4all

Homepage Generator

Related

Questions about Torchscan