AGENTS
337Cursor rule ✓ vendor
../../AGENTS.md.
not rated 1.1k
+12 yesterday A 4 tokens
copy · 100% Apache-2.0
Plugin Gemini CLI ✓ vendor
Agent skills for NVIDIA cuOpt optimization engine: routing, LP/MILP/QP, installation, and server.
not rated 1.1k
+12
changed 23d ago A
tokens not measured
original Apache-2.0
Skill Claude CodeCodex ✓ vendor
Add a new cuTile GPU kernel operator to TileGym. Covers dispatch registration in ops.py, cuTile backend implementation, init.py exports, test creation, and benchmark in tests/benchmark. Use when adding, creating, or implementing a new cuTile operator/kernel in TileGym, or when asking how to register a new cuTile op.
not rated 834
+2 yesterday A SkillSpector: pass 79 tokens
Skill Claude CodeCodex ✓ vendor
Converts cuTile Python GPU kernels (@ct.kernel) to cuTile.jl Julia equivalents. Handles kernel syntax translation, 0-indexed to 1-indexed conversion, broadcasting differences, memory layout (row-major to column-major), type system mapping, and launch API differences. Use when converting, porting, or translating cuTile…
not rated 834
+2
changed 8d ago A SkillSpector: pass 97 tokens
Skill Claude CodeCodex ✓ vendor
Converts cuTile GPU kernels (@ct.kernel) to Triton (@triton.jit). Handles standard in-repo conversion, debugging (cudaErrorIllegalAddress, shape mismatch, numerical mismatch), and mapping cuTile idioms (ct.load/ct.store, ct.Constant, ct.launch) to Triton equivalents. Covers dual-kernel layout flags (e.g.…
not rated 834
+2
changed 21d ago A SkillSpector: pass 125 tokens
Skill Claude CodeCodex ✓ vendor
Use this skill to convert, port, or translate Triton-TileIR or cuTile-Python GPU kernels to cutile-rs (Rust). The orchestrator runs scripts/preflight.sh, then drives a bounded Agent A -> B -> D -> E pipeline (Agent C is diagnostic, Agent F optional), delegating all kernel/host/correctness/perf work to sub-agents and…
not rated 834
+2 yesterday A SkillSpector: warn 103 tokens
Skill Claude CodeCodex ✓ vendor
Use when adding, modifying, optimizing, or debugging CuTile autotuning code. Trigger signals: exhaustivesearch / replacehints / hintsfn / cuda.tile.tune in code, autotune in filenames, or correctness/performance issues in autotuned CuTile kernels. Covers: tune-once/cache/launch pattern, per-architecture configs…
not rated 834
+2 yesterday A SkillSpector: pass 117 tokens
Skill Claude CodeCodex ✓ vendor
Expert cuTile programming assistant. Write high-performance GPU kernels using cuTile's tile-based programming model with proper validation and optimization. Supports deep agent orchestration for complex multi-kernel tasks.
not rated 834
+2
changed 21d ago A SkillSpector: warn 43 tokens
Skill Claude CodeCodex ✓ vendor
Iteratively optimize cuTile kernel performance through systematic profiling, bottleneck analysis, IR comparison, and targeted tuning. Covers tile sizes, occupancy, autotune configs, TMA, latency hints, persistent scheduling, numctas, flushtozero, and IR-level debugging. Use when asked to "optimize cutile kernel"…
not rated 834
+2 yesterday A SkillSpector: pass 111 tokens
Skill Claude CodeCodex ✓ vendor
Integrate TileGym kernels into Hugging Face transformers models by replacing the library's submodule(s) and certain class(es)' implementations, and patching certain class(es)' init/forward/load weight methods prior to instantiating models. Used when the user requires integrating TileGym kernels into transformers…
not rated 834
+2 yesterday A SkillSpector: pass 78 tokens
Skill Claude CodeCodex ✓ vendor
Question-gated router for the curated cuTile optimization wiki: kernel-family playbooks, measured techniques with caveats, performance patterns, and practical language knowledge for Blackwell and Hopper. Use when an agent has inspected the current kernel and has a concrete question that could change its plan, when…
not rated 834
+2 yesterday A 110 tokens
api-caller
348Skill Claude CodeCodex ✓ vendor
Call any REST API dynamically. Make GET, POST, PUT, DELETE requests to any endpoint with custom headers and JSON body.
not rated 643
+26 yesterday A SkillSpector: warn 29 tokens
original Apache-2.0
calculator
349Skill Claude CodeCodex ✓ vendor
Evaluate mathematical expressions and unit conversions. Handles arithmetic, percentages, exponents, and common unit conversions (temperature, distance, weight). No external dependencies.
not rated 643
+26 yesterday A SkillSpector: pass 32 tokens
original Apache-2.0
Skill Claude CodeCodex ✓ vendor
Use when converting an existing benchmark, rubric, verifier, task YAML/JSON, or domain check into SkillEvaluator BYOG/BYOT custom evaluation.
not rated 643
+26 yesterday A SkillSpector: pass 35 tokens
original Apache-2.0
task-list
351Skill Claude CodeCodex ✓ vendor
Required for 4+ step requests; add tasks at start and update status after each step.
not rated 643
+26 yesterday A SkillSpector: pass 22 tokens
original Apache-2.0
text-analyzer
352Skill Claude CodeCodex ✓ vendor
Analyze text content and produce statistics including word count, line count, character count, most frequent words, and readability metrics. Works on any plain text input provided inline or from a file path.
not rated 643
+26 yesterday A SkillSpector: pass 42 tokens
original Apache-2.0
Plugin Claude Code ✓ vendor
Plugin marketplace listing 1 plugin: nvidia-kaggle.
not rated 349
+4 3mo ago A
tokens not measured
original MIT
Skill Codex ✓ vendor
Diagnose a TensorRT-Model-Connect family whose native output disagrees with its declared reference or family-owned correctness contract.
not rated 289
+4
changed 2mo ago A SkillSpector: pass 31 tokens
original Apache-2.0
doc-sync
355Skill Codex ✓ vendor
Audit and update TensorRT-Model-Connect documentation when code, public APIs, family ownership, commands, or model support have changed.
not rated 289
+4
changed 2mo ago A SkillSpector: pass 30 tokens
original Apache-2.0
fp16-trt-network
356Skill Codex ✓ vendor
Implement or review FP16 and BF16 behavior in one family-owned, strongly typed TensorRT network without changing unrelated families.
not rated 289
+4
changed 2mo ago A SkillSpector: pass 30 tokens
original Apache-2.0
Skill Codex ✓ vendor
Evaluate supported precision, quantization, or selected FP32-layer choices for one TensorRT-Model-Connect family with matched correctness and timing.
not rated 289
+4
changed 21d ago A SkillSpector: pass 34 tokens
original Apache-2.0
pr-babysitter
358Skill Codex ✓ vendor
Use when monitoring GitHub pull request CI, diagnosing failed checks, rebasing branches onto github/main, applying narrowly scoped fixes, and updating PRs until their latest checks are green or a human blocker is identified.
not rated 289
+4
changed 1mo ago A SkillSpector: warn 48 tokens
original Apache-2.0
profile-model
359Skill Codex ✓ vendor
Measure one TensorRT-Model-Connect model through its public Task API or run a checked-in performance-matrix entry with comparable evidence.
not rated 289
+4
changed 21d ago A SkillSpector: pass 30 tokens
original Apache-2.0
Skill Claude CodeCodex ✓ vendor
Prepare a TensorRT-Model-Connect development or validation environment from a fresh checkout on a host without a known working setup.
not rated 289
+4
changed 2mo ago A SkillSpector: pass 32 tokens
original Apache-2.0
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: