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/datathings/marketplace/openclnpx skills add datathings/marketplace --skill openclgit clone --depth 1 https://github.com/datathings/marketplaceWrote 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/datathings/marketplace/opencl)<a href="https://agentmods.dev/skills/datathings/marketplace/opencl"><img src="https://agentmods.dev/badge/skills/datathings/marketplace/opencl.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 | $0.00099 | $0.02155 |
| Opus 5 | $0.00049 | $0.01077 |
| Sonnet 5 | $0.00020 | $0.00431 |
| Haiku 4.5 | $0.00010 | $0.00215 |
Grade A, and why
opencl 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 — 132 lines — stays where its author put it; the contents beside it link to each section on GitHub.
OpenCL SDK
Version: v2026.05.29 (Khronos Group OpenCL-SDK) Language: C (OpenCL 1.0–3.1) / C++ (opencl.hpp wrapper) License: Apache-2.0 Repo: https://github.com/KhronosGroup/OpenCL-SDK
Overview
OpenCL (Open Computing Language) is a framework for parallel programming across heterogeneous platforms — GPUs, CPUs, FPGAs, and DSPs — from a single API. The SDK bundles:
- OpenCL-Headers — C headers (
<CL/cl.h>,<CL/cl_ext.h>) - OpenCL-CLHPP — C++ wrapper (
<CL/opencl.hpp>) - OpenCL-ICD-Loader — runtime dispatch to installed platform drivers
- OpenCLUtils / OpenCLSDK — utility libraries (
<CL/Utils/>,<CL/SDK/>)
Quick Start (C)
Kernel file saxpy.cl:
__kernel void saxpy(float a, __global float *x, __global float *y) {
int i = get_global_id(0);
y[i] = fma(a, x[i], y[i]);
}
Host:
#include <CL/cl.h>
cl_platform_id plat; cl_device_id dev;
clGetPlatformIDs(1, &plat, NULL);
clGetDeviceIDs(plat, CL_DEVICE_TYPE_DEFAULT, 1, &dev, NULL);
cl_context ctx = clCreateContext(NULL, 1, &dev, NULL, NULL, &err);
cl_command_queue q = clCreateCommandQueueWithProperties(ctx, dev, NULL, &err);
// ... load source, clCreateProgramWithSource, clBuildProgram,
// clCreateKernel, clSetKernelArg, clEnqueueNDRangeKernel,
// clEnqueueReadBuffer, clReleaseXxx ...
Quick Start (C++)
#define CL_HPP_ENABLE_EXCEPTIONS
#define CL_HPP_TARGET_OPENCL_VERSION 300 // default is now 310 (OpenCL 3.1); minimum 200
#include <CL/opencl.hpp>
cl::Context ctx{CL_DEVICE_TYPE_DEFAULT};
cl::Device dev = ctx.getInfo<CL_CONTEXT_DEVICES>()[0];
cl::CommandQueue queue{ctx, dev};
cl::Program prog{ctx, source_string};
prog.build(dev);
auto saxpy = cl::KernelFunctor<cl_float, cl::Buffer, cl::Buffer>(prog, "saxpy");
saxpy(cl::EnqueueArgs{queue, cl::NDRange{N}}, a, buf_x, buf_y);
Core Concepts
- Work-item — one parallel execution unit; maps to one GPU thread
- Work-group — block of work-items sharing local memory and barriers
- NDRange — N-dimensional index space (up to 3D); defines total parallelism
- Context — owns devices, memory objects, programs, and queues
- Command Queue — ordered or OOO stream of commands to one device
- Memory object — buffer (linear) or image (typed, sampled); device-side
- Kernel — a
__kernelfunction compiled from OpenCL C source or SPIR-V - Event — synchronization token returned by every enqueue command
- Address spaces —
__global(buffers),__local(shared),__constant(read-only),__private(per-item)
What ships with it
7 files 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 · 132 lines · 99 tokens per session scan A b5bfd5daeb0b
opencl is a skill published in the GitHub repository datathings/marketplace (11 stars, last pushed 9d ago), licensed Apache-2.0. It adds 99 tokens to every session and 2,155 once invoked, about $0.0005 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
cuopt-numerical-optimization-api
LP, MILP, and QP (beta) with cuOpt — Python, C, and CLI. Use when the user is solving LP, MILP, or QP with any cuOpt interface.
cuopt-developer
Modify, build, test, debug, and contribute to NVIDIA cuOpt (C++/CUDA, Python, server, CI). Use for solver internals, PRs, DCO, and code conventions.
huawei-cloud-ascendc-operator-performance-optim
Develop and optimize custom operators using AscendC programming language. Analyze operator performance bottlenecks and conduct optimization validation. Based on AscendC and CANN toolkit Use this skill when the user wants to: (1) optimize performance-critical operators on Ascend NPU, (2) develop custom operators for…
cuda-c-basics
CUDA C 核心概念、内核结构和标准编程模式.
torchtalk-analyzer
Analyze PyTorch internals across Python, C++, and CUDA layers using the TorchTalk MCP server. Use when asked about how PyTorch operators work internally, where functions are implemented, what would break if code is modified, or finding tests for PyTorch operators.
torchtalk-trace
Trace a PyTorch function's cross-language binding chain (Python -> C++ -> CUDA).