Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/albumentations-team/albucore/torch-performance-optimizationnpx skills add albumentations-team/albucore --skill torch-performance-optimizationgit clone --depth 1 https://github.com/albumentations-team/albucoreWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/albumentations-team/albucore/torch-performance-optimization)<a href="https://agentmods.dev/skills/albumentations-team/albucore/torch-performance-optimization"><img src="https://agentmods.dev/badge/skills/albumentations-team/albucore/torch-performance-optimization.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00060 | $0.00538 |
| Opus 5 | $0.00030 | $0.00269 |
| Sonnet 5 | $0.00012 | $0.00108 |
| Haiku 4.5 | $0.00006 | $0.00054 |
Grade A, and why
torch-performance-optimization scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 30 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Torch Performance Optimization
Read docs/torch-performance-optimization.md completely before inspecting or editing a Torch path. Also read ../performance-optimization/SKILL.md, its canonical performance guide, and ../albucore-benchmarks/SKILL.md completely.
Workflow
- Establish a correctness baseline and benchmark the existing public path. State container, layout, dtype, shape, parameter, stride, thread, and allocation contracts.
- Audit unnecessary Python work, full-volume passes, conversions, materializations, and output repairs before choosing a Torch operator.
- Profile only to discover candidates. Benchmark every viable implementation end to end, including NumPy↔Torch bridges and layout conversion.
- Compare NumPy, OpenCV, NumKong, StringZilla, and Torch where they share semantics. Treat fused operators, layout,
channels_last_3d, and in-place reuse as hypotheses with their own correctness and performance matrices. - Select a route or threshold only from stable public-path evidence. Preserve rejected candidates in the benchmark report when they clarify a boundary.
- Add correctness tests for each accepted route and update the benchmark evidence, public contract, and canonical guide when a reusable rule or limitation is discovered.
Albucore Contract
- Torch is a required dependency. Assume it is already imported for benchmarks.
- Current public Tensor routes are eager CPU paths that do not record autograd inside the primitive and do not use
torch.compile. Do not add device routes, graph-preserving primitive fallbacks, compilation, or their benchmark candidates. - Tensor layouts are explicit and independent of NumPy layouts. Do not infer
NCHWversusCDHWfrom shape sizes. - Caller-prevalidated 3D routers leave validation outside the hot path. Do not add it back while optimizing.
- A Tensor route must preserve the documented container, layout, border, interpolation, rounding, mutation, and aliasing behavior. Its dtype contract must distinguish supported dtype preservation from any explicit fallback conversion.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 30 lines · 60 tokens per session scan A a6f894d2da0d
torch-performance-optimization is a skill published in the GitHub repository albumentations-team/albucore (123 stars, last pushed 8d ago), licensed MIT. It adds 60 tokens to every session and 538 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
hailo-optimize
Optimize and quantize a parsed Hailo HAR using the Hailo Dataflow Compiler (DFC). Auto-loads a matching ALLS model script from the Model Zoo, prepares calibration data (with the correct normalization alignment), runs optimization on GPU when available, and gates on the post-optimize SNR.
hailo-parse
Parse an ONNX or TensorFlow model into a Hailo HAR using the Hailo Dataflow Compiler (DFC). Use when bringing a new model into the Hailo flow. Auto-discovers parser start/end nodes and normalization from the Model Zoo when the model name matches a known network.
hailo-compile
Compile an optimized Hailo HAR into a HEF binary using the Hailo Dataflow Compiler (DFC). Final stage of the DFC flow — takes the quantized HAR from /hailo-optimize and produces a deployable .hef.
Computer Vision Helper
Assist with image analysis, object detection, and visual AI tasks.
ml-for-aec
Computer vision for buildings, image-to-floorplan, generative ML models, performance prediction, structural analysis ML, energy prediction, natural language to design, and point cloud ML for AEC computational design.
dataset-profiling
Use as the FIRST step of any ML task, before choosing a model, to inspect and understand the actual dataset. Works for a LOCAL dataset (Claude reads the files directly) and for a KAGGLE dataset (Claude cannot read /kaggle/input from your machine, so it emits a small profiling cell you run on Kaggle and paste back, or…