// Code review skill specialized for tt-xla (Python + C++ PJRT plugin for Tenstorrent hardware). Covers C++ memory safety, PJRT API patterns, Python test standards, and project-specific conventions.
Triage one tt-forge-models training test failing with a bfloat16 dtype-mismatch RuntimeError (e.g. "mat1 and mat2 must have the same dtype, but got Float and BFloat16", "'<op>' not implemented for 'BFloat16'"). For cross-dtype operands, attempts a minimal loader fix propagating `dtype_override` into the offending tensor constructor, then re-runs CPU + pytest and updates the YAML (passing -> EXPECTED_PASSING; new failure -> KNOWN_FAILURE_XFAIL). For op-not-implemented (no PyTorch kernel), goes straight to KNOWN_FAILURE_XFAIL with the verbatim error. Updates every training entry sharing the affected loader. Never edits inference YAML or `dynamic_loader.py`.
Triage one tt-forge-models training test stuck at FAILED_FE_COMPILATION with reason "tt-forge-models doesn't implement unpack_forward_output for this model." Inspects the model's forward output, registers a handler or writes a per-loader override, and updates the YAML.
Analyze CI benchmark workflow runs from GitHub Actions for the tt-xla project. Produces a markdown report covering failed jobs (with root-cause error extraction via logs and Glean), successful model performance metrics (samples/sec, TTFT, device perf), perf regressions/improvements vs previous nightly, and the full dependency commit chain (tt-xla, tt-mlir, tt-metal). Use this skill whenever the user wants to analyze a CI run, review nightly benchmark results, investigate CI failures, check benchmark performance from a workflow run, or asks about "latest nightly" results. Also trigger when the user pastes a GitHub Actions run URL or mentions a run ID in the context of performance analysis, or asks about perf regressions.
Use when auditing a TTNN model's IR for missed op fusion opportunities — both direct TTNN fusions (a fused ttnn op already exists) and theoretical fusions (the pattern is a single kernel in torch/triton/cuda)
Analyze a GitHub Actions run and summarize failures
Analyzes, debugs and proposes fixes for graph breaks in PyTorch/XLA model compilation. Use when a model generates more graphs than expected during compilation, the user mentions "graph break", or when debugging excessive graph generation in tt-xla pipelines.
| name | code-reviewer |
| description | Code review skill specialized for tt-xla (Python + C++ PJRT plugin for Tenstorrent hardware). Covers C++ memory safety, PJRT API patterns, Python test standards, and project-specific conventions. |
Specialized code review toolkit for the tt-xla project: a PJRT-based backend that enables JAX and PyTorch/XLA on Tenstorrent AI hardware.
Languages: C++20, Python 3.12
Build: CMake + Ninja, Python setuptools (wheel packaging)
Formatting: clang-format (C++, style from .clang-format), black + isort (Python)
Testing: pytest with custom markers
Logging: loguru (C++), Python stdlib logging
CI: pre-commit hooks (black, clang-format, SPDX copyright, trailing whitespace, isort)
When invoked via /code-reviewer, Claude Code loads this file and the reference documents into context, then applies them to review the code. No external tools, containers, or scripts needed.
references/code_review_checklist.mdreferences/coding_standards.mdreferences/common_antipatterns.mdAll review focus areas, coding standards, and antipatterns are documented in the reference files below. Refer to these as the single source of truth during reviews:
references/code_review_checklist.md — Step-by-step review checklist (C++, Python, CMake, general)references/coding_standards.md — Project coding standards for C++ and Pythonreferences/common_antipatterns.md — Antipatterns specific to tt-xla with wrong/right examples