Predict and analyze LDS access conflicts with an exact rocKE architecture profile and a serializable semantic result. Use when examining lane-to-LDS byte-address mappings, checking broadcasts or conflict multiplicity, comparing supported gfx profiles, or…
Render and explain MMA (MFMA/WMMA) layout visualizations at any scale — the A/B/C tee, register files, logical tiles, derived/final-C, C store distribution, and LDS view. Answers questions from "how does this distribution map a logical tile to registers?" and…
Design and author a GPU kernel with the rocKE tiling API, end-to-end. Dispatches the Tiling Kernel Architect (lead) who consults MMA/LDS/Profiling specialists and drives layout-viz / bank-conflict / build / profile. Any algorithm — the stage classes are…
Create or extend a hipDNN generic-kernel-ingestor integration: graph/kernel contracts, descriptors, native hooks, packaging, host checks, exact-engine device correctness and final corpus coverage.
Run, read, and extend hipDNN's cross-provider integration test suite (dnn-providers/integration-tests): bundles and sweeps, golden data and DVC, .support.json claim sidecars, per-engine TOML, tier YAML, CTest lanes. Doc-driven: loads all of its knowledge at…
Build hipDNN with providers via the repository superbuild. Faster than standalone since providers build alongside hipDNN in a single CMake invocation. On Windows, auto-runs the wheel-based ROCm setup if not already prepared.
Run clang-tidy on hipDNN after a build, scoped to the translation units affected by changed files. Much faster than ENABLE_CLANG_TIDY=ON, which runs clang-tidy inline with every compile. Resolves changed headers to the translation units that include them. Use…
Dispatch TheRock Multi-Arch CI or the hipDNN superbuild CI on a rocm-libraries branch with chosen GPU families, projects and test labels, then check status or watch the run. Always dry-runs first; a real dispatch needs explicit user approval.
Generate hipDNN operation boilerplate from a YAML config. Use when the user wants to add a new operation type to hipDNN, or generate descriptor/packer/unpacker code.
GitHub CI supplements to the ROCm PR quality base skill. Use when authoring, reviewing, or merge-gating a rocm-libraries PR that touches .github/ (workflows, actions, CI scripts). Adds and tightens base rules.
Analyze LDS bank conflicts for a kernel/layout, in one of two explicit modes — investigate (rocprof hardware counters on the host GPU, gated against the simulator) or simulate (a validated per-arch model, no GPU, every number labelled SIMULATED). Use whenever…
hipBLASLt supplements to the ROCm PR quality base skill. Use for hipBLASLt PR author, review, or pre-merge gating (target branch develop; product paths under projects/hipblaslt/**, including tensilelite/). Adds and tightens base rules; never relaxes a base…
Guide optional, offline TensileLite mutation analysis used to find behavioral coverage and assertion gaps that should be closed with focused characterization tests. Use when planning, running, resuming, or auditing a TensileLite mutmut campaign; inspecting a…
hipDNN supplements to the ROCm PR quality base skill. Use for hipDNN PR author, review, or pre-merge gating (target branch develop; product paths under projects/hipdnn/** and the providers under dnn-providers/**). Adds and tightens base rules; never relaxes a…
Locate where a feature lives in the StinkyTofu codebase, and advise where and how to add a new one, grounded in the actual source. Use when the user asks "where is X", "where should this new feature go", "how do I add a pass / instruction / architecture /…
Give each git worktree its own Python venv for any rocm-libraries package (rocisa, stinkytofu bindings, ...) without duplicating a multi-GB pip ROCm SDK. Use when concurrent worktrees fight over one shared venv, an editable install from one worktree shadows…
DEPRECATED — superseded by hipdnn-pr-quality. Reviewing hipDNN PRs and diffs is now the review-assist action of the hipdnn-pr-quality overlay. This stub remains only so the old name still resolves and redirects.
DEPRECATED — superseded by hipdnn-pr-quality. Drafting PR titles and bodies is now the author-assist action of the hipdnn-pr-quality overlay. This stub remains only so the old name still resolves and redirects.
Turn a technical RFC into an actionable backlog of JIRA stories and tasks, grounded in the codebase the RFC affects. Derives unified, independently-shippable work items (each with a user-story sentence, acceptance criteria, functional and non-functional…
Deep-dive review of a technical RFC focused on change-management concerns — API/ABI/wire/on-disk compatibility, version skew, migration tooling, rollout mechanics, reversibility, and consumer impact. Produces a written review report. Use when an RFC touches a…
Deep operational-concerns review of a technical RFC — build system impact, packaging, CI, observability, performance-in-production, deployment, capacity, and failure modes. Use as a focused follow-up to /rfc-review when the operational lens deserves more than…
Deep-dive security review of a technical RFC or design document — threat model, trust boundaries, attack surface, input validation, crypto, authN/authZ, secrets, supply chain, privacy, and audit. Use when reviewing an RFC and security warrants its own focused…
Review a technical RFC or design document across a structured set of perspectives (framing, design correctness, change management, operational concerns, sustainability) and produce a written review report. Use when reviewing RFCs, design docs, ADRs, or…