Your agent can do more. Find the skill that makes it happen.
Find your next superpower If you are an agent, refer to our llms.txt for full access.Stores and retrieves genomic variant calls with TileDB-VCF. Use for indexed single-sample VCF/BCF ingestion, incremental cohorts, region and sample queries, streaming results, allele statistics, QC, and VCF/BCF export locally or through TileDB Cloud.
Covers registering new operators in ops.py, implementing the cuTile backend, and exposing exports through __init__.py. Includes test creation and benchmarks in tests/benchmark.
Translates kernel syntax, indexing, broadcasting, memory layout, types, and launch APIs from Python cuTile to Julia cuTile.jl. Also supports debugging and optimizing existing Julia translations.
Maps cuTile kernel operations, constants, and launch patterns to Triton equivalents. Covers in-repository conversions, dual-kernel layout flags, autotuning grids, and debugging illegal memory access, shape mismatches, and numerical discrepancies.
Covers tune-once, cache, and launch patterns, parameter space design, and configurations for sm80 through sm120. Addresses tile sizes, occupancy, num_ctas, and common correctness or performance pitfalls.
Provides assistance with high-performance GPU kernels using cuTile's tile-based programming model. Includes validation, optimization, and deep agent orchestration for complex tasks involving multiple kernels.
Uses profiling, IR comparison, benchmarking, and targeted tuning to improve GPU kernel performance. Covers tile sizes, occupancy, autotuning, TMA, latency hints, persistent scheduling, num_ctas, flush_to_zero, and IR-level debugging.
Replaces selected transformers submodules and class implementations with TileGym kernels. Patches initialization, forward, and weight-loading methods before model instantiation.
Creates a new Tile Palette asset for organizing tiles. Supports user-specified grid layouts such as rectangular, hexagonal, and isometric.
Creates empty RuleTile, HexagonalRuleTile, or IsometricRuleTile assets. Intended for custom rule configuration when no sprites or sprite sheets are provided or referenced.
Generates a narrative development-history report from a project's entire claude-mem timeline. Supports project journey reports, timeline analysis, and full historical project reports.
Performs zero-shot time-series forecasting with Google's TimesFM, including regular-grid CSV preparation, quantile forecasts, XReg covariates, and held-out evaluation. Uses the Apache-licensed TimesFM 2.5 checkpoint by default and documents the distinct TimesFM 3.0 multivariate API and weight-license requirements.
Skills give your agent reusable instructions for a specific job. Pick one, read what it does, and bring it into your workflow.
A name only tells half the story. Search the full description and instructions to find the right fit.
Read the skill, visit its source, and see exactly what you’re adding to your agent.
Copy the install command from a skill page and run it in your project.
npx skillycli add owner/repo --skill name