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 skills add NVIDIA/TileGym --skill tilegym-converting-cutile-triton-to-cutile-rsgit clone --depth 1 https://github.com/NVIDIA/TileGymWrote 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/nvidia/tilegym/tilegym-converting-cutile-triton-to-cutile-rs)<a href="https://agentmods.dev/skills/nvidia/tilegym/tilegym-converting-cutile-triton-to-cutile-rs"><img src="https://agentmods.dev/badge/skills/nvidia/tilegym/tilegym-converting-cutile-triton-to-cutile-rs/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/nvidia/tilegym/tilegym-converting-cutile-triton-to-cutile-rs"><img src="https://agentmods.dev/badge/skills/nvidia/tilegym/tilegym-converting-cutile-triton-to-cutile-rs.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Rogue Agent · line 140 Skill modifies its own code, configuration, or behavior at runtime. Self-modification enables an agent to escalate privileges, disable safety constraints, or install persistent backdoors.Fix: Prevent the skill from modifying its own code, SKILL.md, or configuration files. Treat skill files as read-only at runtime.
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 | $0.00103 | $0.04792 |
| Opus 5 | $0.00051 | $0.02396 |
| Sonnet 5 | $0.00021 | $0.00958 |
| Haiku 4.5 | $0.00010 | $0.00479 |
Grade A, and why
tilegym-converting-cutile-triton-to-cutile-rs 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 13d 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.
The source is not reproduced here
A licence we could not identify
The repository carries a LICENSE file, but it is custom or dual enough that GitHub cannot name it and neither can this catalogue. Unknown terms are not permission, so the body is not copied here. Read the licence at the source and decide for yourself.
What ships with it
43 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.
- agents/agent_a.md 15 KB
- agents/agent_b.md 24 KB
- agents/agent_c.md 11 KB
- agents/agent_d.md 29 KB
- agents/agent_e.md 17 KB
- agents/agent_f.md 9.6 KB
- BENCHMARK.md 4.6 KB
- concepts/ffi-bridge.md 11 KB
- concepts/strided-view-to-partition-view.md 6.7 KB
- concepts/tensor-vs-pointer-pattern.md 4.4 KB
- concepts/transpose-support.md 8.7 KB
- evals/evals.json 4.8 KB
- examples/bmm/ffi.rs 7.0 KB
- examples/bmm/kernel.rs 8.5 KB
- examples/bmm/walkthrough.md 5.4 KB
- examples/bmm/wrapper.py 6.9 KB runs code
- examples/softmax/ffi.rs 4.5 KB
- examples/softmax/kernel.rs 9.4 KB
- examples/softmax/softmax_pipeline.rs 6.1 KB
- examples/softmax/walkthrough.md 12 KB
- examples/softmax/wrapper.py 18 KB runs code
- references/coding-rules.md 39 KB
- references/env-block.md 3.3 KB
- references/ir-diff-checklist.md 2.6 KB
- references/no-mut-tensor-output.md 4.0 KB
- references/op-mapping.md 14 KB
- references/output-structure.md 4.5 KB
- references/performance-checklist.md 11 KB
- references/pipeline.md 16 KB
- scripts/diff_ir.sh 15 KB runs code
- scripts/preflight.sh 6.3 KB runs code
- scripts/validate_agent_b.sh 8.3 KB runs code
- scripts/validate_agent_d.sh 14 KB runs code
- scripts/validate_agent_e.sh 2.5 KB runs code
- scripts/validate_agent_f.sh 1.4 KB runs code
- scripts/validate_analysis.sh 11 KB runs code
- scripts/validate_diff_report.sh 5.6 KB runs code
- scripts/validate_entry_pattern.sh 10 KB runs code
- scripts/validate_ir.sh 3.0 KB runs code
- scripts/validate_kernel.sh 19 KB runs code
- scripts/validate_perf_log.sh 1.6 KB runs code
- skill-card.md 4.3 KB
- skill.oms.sig 14 KB
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.
- 13d ago First seen · 309 lines · 103 tokens per session scan A a4e9a626ebef
tilegym-converting-cutile-triton-to-cutile-rs is a skill published in the GitHub repository NVIDIA/TileGym (808 stars, last pushed yesterday), with no licence file. It adds 103 tokens to every session and 4,792 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
rust-patterns
Idiomatic Rust patterns, ownership, error handling, traits, concurrency, and best practices for building safe, performant applications.
dd-code-generation
Use pup CLI for immediate Datadog operations or generate code for integration into applications.
omh-rust
This is a Hermes-native rust workflow skill.
bevy-ecs
Structure a Bevy app around its Entity Component System: build the App with plugins, define Component/Resource types, write systems with Query/Res/Commands, filter and order systems, and use the Time resource for frame-rate-independent motion. Use when building or debugging a Bevy game in Rust — when the user mentions…
rust-project
Modern Rust project architecture guide for 2025. Use when creating Rust projects (CLI, web services, libraries). Covers workspace structure, error handling, async patterns, and idiomatic Rust best practices.
azure-keyvault-secrets-rust
Client library for Azure Key Vault Secrets — secure storage for passwords, API keys, and other secrets.