all publishers

tenstorrent

@tenstorrent source repo

88 published skills

  1. ▌
    Feature Branch Pr 3 · tenstorrent
    Make a change in tt-model-manager the team way: branch off `main` with a `fix/<slug>` / `feat/<slug>` (or `<username>/<slug>`) branch, keep the diff minimal and single-concern, add a regression test, run the full offline pytest suite before pushing, leave every other branch untouched, and open a **draft** PR against `main` with a human, professional title and description. Use whenever the user asks to start a feature/fix/branch, make a change and open a pull request, or "do this properly on a new branch" in this repo. Never pushes to `main` and never merges its own PR.
    125 repo stars
  2. ▌
    Perf Parameter Impact 2 · tenstorrent
    Analyze LLK performance-report CSV files and create evidence-backed parameter-impact reports in a Cursor Canvas. Use when asked how formats, fidelity, dimensions, accumulation, indexing, loop factors, or other parameters affect TILE_LOOP or another marker in a .post.csv report.
    125 repo stars
  3. ▌
    Arch Lookup 2 · tenstorrent
    Look up Tensix HW behavior across architectures specifically to inform tt-emule mock implementations. Orchestrates per-arch sage agents in parallel and aggregates results with emule-mapping context.
    125 repo stars
  4. ▌
    Feature Branch Pr 2 · tenstorrent
    Make a change in tt-cli the team way: branch off `main` with a `<username>/<feature>` branch, keep the diff minimal and in-scope, verify with the fake-mode test suite (and the real CLI where it applies) before pushing, leave every other branch untouched, and open a PR against `main` with a human, professional title and description. Use whenever the user asks to start a feature/fix/branch, make a change and open a pull request, or "do this properly on a new branch" in this repo. Enforces no AI-tool attribution in commits, PR text, or review comments.
    125 repo stars
  5. ▌
    Agentic Workflows 2 · tenstorrent
    Route gh-aw workflow design/create/debug/upgrade requests to the right prompts.
    125 repo stars
  6. ▌
    CLI Design 2 · tenstorrent bundle
    The complete CLI design language distilled from TT-Studio's `python run.py`, made portable to any repo: the command surface (one self-bootstrapping entrypoint, grouped `--help` panels, early-dispatch utility flags, lifecycle symmetry, re-viewable state), the run's spine (a fixed phase stepper, one collapsing line per step, minimal-by-default folding with `-v`), progress parsed out of subprocess streams, failures rendered as diagnosis cards instead of log dumps, lenient defaults with opt-in strictness, and the testing/verification patterns that keep it honest. Use when building a setup/launcher/dev-tool CLI, adopting this style in another repo, or cleaning up a CLI that dumps raw tool output, tracebacks, or ungrouped flags — and whenever someone says the CLI "looks bad". Ships runnable reference files (`cli_skeleton.py`, `console.py`, `demo.py`, tests) plus a one-command `install.py`.
    125 repo stars
  7. ▌
    Agentic Workflows · tenstorrent
    Route gh-aw workflow design/create/debug/upgrade requests to the right prompts.
    125 repo stars
  8. ▌
    Perf Report 2 · tenstorrent
    Generate, refresh, and validate an LLK performance report by running a perf sweep end to end and checking the resulting perf_data CSV. Use when asked to produce or refresh a perf report for an op, run a perf sweep, or when a report is missing, partial, stale, or implausible.
    125 repo stars
  9. ▌
    Quasar Perf Test 2 · tenstorrent
    Create, extend, debug, and validate Quasar LLK performance tests, their PerfRunType kernel paths, and tile/dimension sweep coverage. Use when adding perf_[op]_quasar.py, wiring PerfConfig, implementing UNPACK_ISOLATE / MATH_ISOLATE / PACK_ISOLATE / L1_CONGESTION, choosing PERF_TILE_SIZES or dest-fill matrices, or investigating implausible Quasar perf metrics and dvalid handshakes.
    125 repo stars
  10. ▌
    Add Project · tenstorrent
    Guide the user through submitting a new entry to the Awesome Tenstorrent list. Use when the user wants to add, submit, or propose a project, library, tool, or resource to tt-awesome.
    125 repo stars
  11. ▌
    Perf Benchmark Single Chip · tenstorrent bundle
    Run device performance benchmarks for tt-blacksmith single-chip training workloads. Use when the user asks to benchmark, profile, or measure performance of a training workload on Tenstorrent hardware, or mentions tracy, tt-perf-report, or device time analysis.
    125 repo stars
  12. ▌
    Llk Test Runner Skill · tenstorrent
    Delegates LLK test runs to the llk-test-runner agent using @.cursor/agents/llk-test-runner.md. Use when the user asks to run tests or mentions LLK tests. Ensure test commands run to completion before reading terminal output (no polling).
    125 repo stars
  13. ▌
    Extend Workflow Engine · tenstorrent
    Checklist for adding or modifying code in the workflow engine (run_workflows.py plus the llm_module/report_module/test_module/workflow_module packages) — new models, workflows, media runners, spec tests, report kinds, or CLI flags — and wiring it back to the entry point run.py through workflows/workflow_dispatch.py. Use whenever code lands in those modules, when onboarding a model type to the engine, or when an engine feature "works standalone via run_workflows.py but not through run.py".
    125 repo stars
  14. ▌
    Add Model Dynamo · tenstorrent bundle
    Checklist for onboarding a new LLM to the cpp_server inference backend so it serves through Dynamo — registering the model type, fetching tokenizer files, tokenizer static data (eos/stop/think tokens), and Dynamo discovery (reasoning + tool-call parsers, generation_config publishing). Use whenever a new model is being onboarded to cpp_server, a HuggingFace model is being wired into the Dynamo deploy, or a model "loads but generates wrong / isn't discoverable / isn't selectable".
    125 repo stars
  15. ▌
    Build Blaze Images · tenstorrent
    Trigger Tenstorrent Blaze media server and tt-metal upstream Docker image builds with gh. Use when the user asks to build Blaze and Metal Docker images, run the tt-shield media server image workflow, or build tt-metal images from the tt-llm-engine submodule reference.
    125 repo stars
  16. ▌
    Profile Cpp Server · tenstorrent
    Capture on-CPU and off-CPU flamegraphs of the running tt_media_server_cpp main (Drogon) and worker processes using Linux perf + Brendan Gregg's FlameGraph. Use when the user asks to profile the C++ server, find a performance bottleneck, identify slow code paths, capture a flamegraph, or investigate CPU usage / lock contention in BlazeRunner or the decode scheduler.
    125 repo stars
  17. ▌
    CLI Design · tenstorrent bundle
    The complete CLI design language distilled from TT-Studio's `python run.py`, made portable to any repo: the command surface (one self-bootstrapping entrypoint, grouped `--help` panels, early-dispatch utility flags, lifecycle symmetry, re-viewable state), the run's spine (a fixed phase stepper, one collapsing line per step, minimal-by-default folding with `-v`), progress parsed out of subprocess streams, failures rendered as diagnosis cards instead of log dumps, lenient defaults with opt-in strictness, and the testing/verification patterns that keep it honest. Use when building a setup/launcher/dev-tool CLI, adopting this style in another repo, or cleaning up a CLI that dumps raw tool output, tracebacks, or ungrouped flags — and whenever someone says the CLI "looks bad". Ships runnable reference files (`cli_skeleton.py`, `console.py`, `demo.py`, tests) plus a one-command `install.py`.
    125 repo stars
  18. ▌
    Tt Model YAML · tenstorrent
    Author a tt-model.yaml — the CONTAINER (v5.1) manifest only — for a model in a tt-metal checkout, by reading its serve recipe and import closure and interviewing the user for what the directory cannot tell you. Use for "containerize this model", "write a tt-model.yaml", "package this as a Docker image", or "make a v5.1/container package". Do NOT use for a v5 "fat" bundle (`tt-model package --from-metal --ttnn-wheel …`) or a v6 "thin" bundle (`tt-model package-thin …`) — those are authored with CLI flags and have no manifest file; this skill would produce a YAML they cannot read.
    125 repo stars
  19. ▌
    Tt Model Package Test · tenstorrent
    Package a model as a CONTAINER (v5.1) image from its tt-model.yaml and prove it works — build, serve on hardware, verify the API (tool calling included), stop cleanly, and optionally push. Use for "package this model", "build and test the container", "prove the package serves", "ship this model to HF". Requires an authored tt-model.yaml (use the tt-model-yaml skill to write one) and a box with the target TT devices. Do NOT use for v5 fat / v6 thin bundles — see AGENTS.md "Driving the flow" for those.
    125 repo stars
  20. ▌
    Model Bringup · tenstorrent
    Bring up a model that tt-inference-server lists as supported but that TT-Studio can't deploy today — add it to the catalog, deploy it on real hardware, triage failures layer by layer (tt-studio → tt-inference-server → tt-metal), verify fixes by hot-patching the running container, ship a patched GHCR image when upstream fixes can't wait, and open minimal PRs in every repo at fault. Use when asked to "add model X for <board>", "bring up <model> from the inference server", or when a model_support doc URL from tt-inference-server is shared. Ends with a tt-studio PR opened via the feature-branch-pr skill.
    125 repo stars
  21. ▌
    Tt Studio CLI · tenstorrent
    The TT-Studio CLI skill — how the `python run.py` command-line interface (the `tt_setup/` package) is built and how to change it the TT-Studio way: Typer CLI conventions (flag grouping, early-dispatch utility flags, `--help`/`--help-env` docs), package layout and the re-export pattern, the calm minimal-by-default terminal output, hardware/QB2 detection (IS_QB2 opt-in, verify-don't-trust), and the testing/verification patterns (pure helpers, PTY render checks). Use for any CLI work: BEFORE adding or changing a `python run.py` flag or its help text, editing anything under `tt_setup/` or its terminal output, touching startup/teardown flow, or changing hardware detection.
    125 repo stars
  22. ▌
    Feature Branch Pr · tenstorrent
    Make a change in TT-Studio the team way: branch off `dev` with a `<username>/<feature>` branch, keep the diff minimal and in-scope, verify the running app via its health endpoints before pushing, leave every other branch untouched, and open a PR against `dev` with a human, professional title and description. Use whenever the user asks to start a feature/fix/branch, make a change and open a pull request, or "do this properly on a new branch" in this repo. Enforces no AI-tool attribution in commits, PR text, or review comments.
    125 repo stars
  23. ▌
    Tt Studio Overview · tenstorrent
    Project map for TT-Studio — what the platform is, where its docs live, its key components (React frontend, Django backend, FastAPI inference server, Docker), and the AI model types it supports. Use when you need a high-level orientation of the repo, want to know which doc covers a topic, or are trying to locate where a component or capability lives before diving into code.
    125 repo stars
  24. ▌
    Draft Release Notes · tenstorrent
    Draft TT-Studio GitHub release notes in the house style: study the last one or two published releases for structure and voice, diff the current release branch against the previous release tag/commit, then group the changes into 🚀 New Features / 🐛 Bug Fixes / 🔧 Technical Improvements with bold feature names and short descriptions. Use whenever someone asks to "draft a release note", "write release notes for vX.Y.Z", or "make release notes from the rc-vX.Y.Z branch". Delivers the final result as a copyable markdown code block.
    125 repo stars
  25. ▌
    Tt Studio Debug Bundle · tenstorrent
    Triage a TT-Studio bug-report ZIP / log bundle (named `tt-studio-logs-ttbr-*` or referenced as `ttbr-*`) to root-cause why a model deploy, board allocation, or Voice Agent pipeline failed. Use when the user shares a path to a TT-Studio log bundle, mentions a bug report ID like `ttbr-<hex>`, or attaches a screenshot of the TT-Studio Bug Report issue template. Fans out parallel Explore subagents across `app/backend/board_control/`, `app/backend/docker_control/`, `app/backend/shared_config/`, and `app/frontend/src/components/` to map log evidence to source code.
    125 repo stars
  26. ▌
    License Attribution Compliance · tenstorrent
    Review third-party license attribution for TT-Studio when dependencies or bundled assets change, and keep the attribution files in sync. Use when a PR adds/updates a dependency (app/backend/requirements.txt, app/agent, inference-api, docker-control-service, or app/frontend/package.json), bundles a binary/model/weights file, or when the user asks to "attribute a license", "update third-party-licenses", "check license compliance", or run the license check. Pairs with the deterministic gate dev-tools/check_license_attribution.py (freshness + new-dependency attribution) that runs on pull requests; this skill does the judgment the script cannot -- classifying licenses and flagging NonCommercial/copyleft/provenance risks.
    125 repo stars
  27. ▌
    Ttnn · tenstorrent bundle
    TTNN operations library reference for Tenstorrent hardware. Covers tensor APIs, ops catalog, model conversion from PyTorch, and memory/layout configuration.
    125 repo stars
  28. ▌
    Tt Lang · tenstorrent bundle
    Comprehensive TT-Lang DSL reference including programming model, APIs, hardware constraints, and guides for translating CUDA, Triton, PyTorch, or TTNN kernels
    125 repo stars
  29. ▌
    Tt Bug Report · tenstorrent
    File a bug report with a reproducer against Tenstorrent repos (tt-lang, tt-metal, tt-xla)
    125 repo stars
  30. ▌
    Tt Enable Tracing · tenstorrent bundle
    TTNN trace capture and replay for eliminating dispatch overhead. Essential for real-time inference and multi-chip performance.
    125 repo stars
  31. ▌
    Tt Lang To Tt Metal · tenstorrent
    Export TT-Lang kernel to TT-Metal C++ kernel code
    125 repo stars
  32. ▌
    Tt Connect Remote Device · tenstorrent bundle
    Set up and verify remote connection to Tenstorrent hardware. Provides tools for running kernels, copying files, and reading logs on remote devices.
    125 repo stars
  33. ▌
    Tt Lang Profile Optimize · tenstorrent bundle
    Profile and optimize TT-Lang kernels for performance. Covers auto-profiling, perf summary, signposts, and optimization workflow.
    125 repo stars
  34. ▌
    Add Op · tenstorrent bundle
    How to add a new operation (op) to the tt-mlir compiler across all layers: TTIR/TTNN dialect definitions, StableHLO composite conversion, TTIR-to-TTNN conversion, EmitC/EmitPy conversions, flatbuffer schema and serialization, runtime implementation, OpModel, ttir_builder, golden functions, and all associated tests. Use this skill whenever the user asks to add an op, implement an op, create a new operation, add support for a TTNN op, or mentions adding an op to the compiler pipeline. Also trigger when the user wants to know what files to change for a new op, or asks about the op-adding workflow.
    125 repo stars
  35. ▌
    Uplift Ttsim CI · tenstorrent
    Uplift the TTSim version used by tt-mlir CI and refresh WH/BH simulator skips. Use when updating `.github/workflows/call-test-ttsim.yml`, changing `ttsim-version`, validating WH or BH TTSim golden tests, or triaging simulator-specific pytest skips.
    125 repo stars
  36. ▌
    Add Ttir Builder Op · tenstorrent
    Add full builder API support (@tag, @parse, @split) for a TTIR op. Use this skill whenever the user wants to add builder support for a new TTIR op, upgrade an existing _op_proxy-based op to use @tag/@parse/@split decorators, or asks about how to add builder API for an op in ttir_builder.py. Also trigger when the user mentions adding tag/parse/split for an op, or wants to make an op work with the parse/split test infrastructure.
    125 repo stars
  37. ▌
    Add Ttir D2m Lowering · tenstorrent
    Elementwise TTIR→D2M→TTMetal path: tablegen, TTIRToD2M.cpp, D2MToTTKernel.cpp, and — only when the kernel API callee is new — TTKernelIncludesMap.h (per-op api/compute/eltwise_unary/*.h mapping for JIT). Does not edit D2MGenericRegionOps.cpp or TTKernelToCpp.cpp. Not for reductions, matmul, views, or CCL.
    125 repo stars
  38. ▌
    Ttir Model Op Analysis · tenstorrent
    Given a `.mlir` file (or a directory of `.mlir` files) with TTIR ops, run the same TTIR normalization passes as `D2MFrontendPipeline` before D2M, then produce per-file outputs: `preprocessed.mlir`, `ttir-op-report.txt` (op counts from normalized IR), and `ops.mlir` (one func per unique op configuration, golden-style). Optional: per-pass IR dumps.
    125 repo stars
  39. ▌
    Run Ops Mlir Snippets · tenstorrent
    Compile and optionally execute every func.func in an ops.mlir-style snippet file (or every .mlir file in a directory) using `run_ops_mlir_snippets.py`. Use when the user wants to compile or run TTIR op snippets on device, test ops.mlir files, or check which ops compile/execute successfully.
    125 repo stars
  40. ▌
    Ttir Decomposition For Ttmetal · tenstorrent
    Add a new composite op decomposition pattern to the TTMetal pipeline. Use when the user wants to decompose/lower a high-level TTIR op (e.g. rms_norm, sdpa, layer_norm, softmax) into primitive TTIR ops (matmul, add, multiply, etc.) for the D2M/TTMetal backend. Also trigger when the user mentions "decomposition pattern", "decompose op for ttmetal", or "lower op to primitives".
    125 repo stars
  41. ▌
    Validate Tt Mlir Against Tt Xla · tenstorrent
    Validate a tt-mlir PR against tt-xla by creating a cherry-picked branch and triggering CI. Invoked as: /validate-tt-mlir-against-tt-xla <PR number or URL>. Use this skill whenever the user wants to test, validate, qualify, or check a tt-mlir PR in tt-xla, or mentions running uplift qualification test suite, or asks to trigger tt-xla CI for a tt-mlir change. Also triggers when the user mentions "xla validate", "xla test", or "validate in xla".
    125 repo stars
  42. ▌
    Uplift · tenstorrent
    Diagnose and fix emule regressions caused by bumping the tt-metal pin (and its bundled tt-umd submodule). Use when a pin bump turns the C++ or TTNN regression red — device-open crashes, JIT-compile errors, hangs, or new data mismatches — and you need to prove the cause and land a faithful emule-side fix.
    125 repo stars
  43. ▌
    Verify Mock · tenstorrent
    Verification checklist to walk before declaring a tt-emule mock complete. Use after implementing a stub (typically at the end of /implement-mock or /compute-llk-bringup) to catch the recurring failure modes — signature drift, no-op math, missing format dispatch, JIT-cache staleness, regression baseline drift.
    125 repo stars
  44. ▌
    Workarounds · tenstorrent
    Registry of deliberate, non-ideal workarounds currently live in tt-emule (hacks we accepted to make something pass, NOT normal development practice). Read this before adding any new workaround, and when touching code that carries one — each entry records what the hack is, why it is not the right fix, the real root cause, and the path to removing it.
    125 repo stars
  45. ▌
    Memory Debug · tenstorrent
    Diagnose data-corruption failures in tt-emule operations (wrong values, partial zeros, off-by-N row/col writes). Use when a regression test reports ATOL > 1e-3 on otherwise-functional kernel paths.
    125 repo stars
  46. ▌
    Implement Mock · tenstorrent
    Workflow for adding a faithful mock of a silicon API to tt-emule. Composes `/arch-lookup` (find HW spec) → emule-mapping strategy selection → stub implementation → verification against sentinel + tt-metal regression.
    125 repo stars
  47. ▌
    Index Based Ops · tenstorrent
    Bring up or debug index-based compute ops (TopK, Sort, Argmax) in tt-emule — ops that return values + indices, where ties make the index choice non-unique. Use when implementing one of these ops or triaging its value/index test failures.
    125 repo stars
  48. ▌
    Shepherd Emule Pr · tenstorrent
    Take an open tt-emule PR to a mergeable, ethos-compliant state. Use when asked to review a PR, address its comments, and prepare it to land — verifies every review point by build (not by trusting "addressed"), runs an emule-ethos pass for the bug classes reviewers miss, then rebases onto main locally and hands the developer a change+verification summary (with a drafted per-point comment and the exact push command) to review and push. The skill never force-pushes the author's branch itself — the human stays in the loop for that.
    125 repo stars
  49. ▌
    Compute Llk Bringup · tenstorrent
    Specialization of `/implement-mock` for compute-kernel LLK shims (`<op>_tile` / `<op>_tile_init` under `include/jit_hw/api/compute/`). Use when bringing up a missing compute op or addressing its PCC failures.
    125 repo stars
  50. ▌
    Parallel Mock Implementation · tenstorrent
    Pattern for parallelizing emule mock authorship via the Workflow tool. Spawn one sub-agent per file in isolation; the orchestrator handles centralized finalization (wiring, build, test, promote). Invoked from `/implement-mock` (Strategy A batching) and `/compute-llk-bringup` (shim sweeps).
    125 repo stars
  51. ▌
    Docs Review · tenstorrent
    Review and fix language correctness and formatting in documentation files.
    125 repo stars
  52. ▌
    Code Reviewer · tenstorrent bundle
    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.
    125 repo stars
  53. ▌
    Triage Inputs · tenstorrent
    Triage missing-input training failures
    125 repo stars
  54. ▌
    Analyze Nightly · tenstorrent
    Analyze a GitHub Actions run and summarize failures
    125 repo stars
  55. ▌
    Compare Nightly · tenstorrent bundle
    Compare two nightly GitHub Actions runs and classify failures as new, persisting, or fixed
    125 repo stars
  56. ▌
    Pull Main Summary · tenstorrent
    Stash local changes, pull latest main, and display a structured commit summary. For tt_forge_models uplift commits, shows the affected models from the submodule log.
    125 repo stars
  57. ▌
    Excalidraw Diagram · tenstorrent bundle
    Create Excalidraw diagram JSON files that make visual arguments. Use when the user wants to visualize workflows, architectures, or concepts.
    125 repo stars
  58. ▌
    Graph Break Analysis · tenstorrent
    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.
    125 repo stars
  59. ▌
    CI Benchmark Analyzer · tenstorrent bundle
    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.
    125 repo stars
  60. ▌
    Triage Dtype Bfloat16 · tenstorrent
    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`.
    125 repo stars
  61. ▌
    Finding Missed Fusions · tenstorrent
    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)
    125 repo stars
  62. ▌
    Sharding Model Analysis · tenstorrent bundle
    Analyze a PyTorch model, then design and implement an optimal multi-device sharding strategy for Tenstorrent hardware that minimizes collective-communication (CCL) ops. A strategy is both which parallelism to use (tensor / sequence / data) and how to shard under it - e.g. Megatron column→row is one of several tensor-parallel schemes. Use whenever the user asks to shard a model, distribute it across devices, write Shardy annotations, reduce CCLs, or analyze a model for tensor/sequence parallelism. Trigger even when the user only says "sharding strategy" or names a model with a device count, without asking for a full plan.
    125 repo stars
  63. ▌
    Transformers Uplift Fix · tenstorrent
    Fix transformers compatibility regressions in tt-xla and tt-forge-models after the pinned version was bumped. Called by the transformers-uplift CI orchestrator with a scope (api-check, model-test-uplifts, model-perf-uplift) and a captured failure context. Edits source-level only — no monkey-patching, no shims, no git operations.
    125 repo stars
  64. ▌
    Triage Unpack Forward Output · tenstorrent bundle
    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.
    125 repo stars
  65. ▌
    Bash Safety · tenstorrent bundle
    Enforce safe bash scripting practices when writing, reviewing, or fixing shell scripts. Covers quoting, arrays, conditionals, arithmetic, redirections, strict mode, and static analysis. Use when editing .sh/.bash files, reviewing shell scripts, fixing shellcheck warnings, or writing new bash code.
    125 repo stars
  66. ▌
    Perturb · tenstorrent
    Reproduce a flaky/timing-dependent kernel failure (suspected undocumented HW race) by injecting NOPs/delays to shift inter-thread / inter-kernel / inter-core timing until the failure becomes frequent enough to isolate and minimize into a deterministic reproducer. Sweeps NOP count × injection position × actor, records the max-error scenario PER actor to a report file (rewritten after each actor finishes so you can stop early), then (on request) minimizes the test around a chosen scenario. Works for tt-metal ttnn op tests and tt-llk kernel tests.
    125 repo stars
  67. ▌
    Run Test · tenstorrent
    Run LLK tests using the test runner agent. Never run pytest directly.
    125 repo stars
  68. ▌
    Arch Lookup · tenstorrent
    Look up Tensix architecture, instruction, or LLK implementation details across architectures. Orchestrates sage agents in parallel.
    125 repo stars
  69. ▌
    Perf Report · tenstorrent
    Generate, refresh, and validate an LLK performance report by running a perf sweep end to end and checking the resulting perf_data CSV. Use when asked to produce or refresh a perf report for an op, run a perf sweep, or when a report is missing, partial, stale, or implausible.
    125 repo stars
  70. ▌
    Port Kernel · tenstorrent
    Get structured porting guidance when moving a kernel between architectures. Launches sages for source and target, reads test harness.
    125 repo stars
  71. ▌
    Debug Kernel · tenstorrent
    Debug compilation or runtime errors in LLK kernels. Infers architecture and kernel type from path.
    125 repo stars
  72. ▌
    Tensor Shape · tenstorrent
    Convert LLK lib/API tile-size args to ckernel::TensorShape and maintain TRISC TensorShape coverage. Use when adding TensorShape parameters, replacing face_r_dim/num_faces, editing LLK_VALIDATE_TENSOR_SHAPE_*, regenerating tensor_shape_coverage_*.h, or reviewing TensorShape PRs.
    125 repo stars
  73. ▌
    Llk Wave Debug · tenstorrent
    Inspect VCS/FSDB waveforms to localize a Quasar LLK runtime hang, timeout, or mismatch when log/source debugging cannot pin the boundary. Use only after the simulator reaches device-ready; not for compile, pre-device-ready simulator, or confirmed environment failures. Quasar only.
    125 repo stars
  74. ▌
    Noc Sync Audit · tenstorrent
    Noc Sync Audit
    125 repo stars
  75. ▌
    Race Audit All · tenstorrent
    Run all nine LLK hazard audits (mmio-race, reconfig-stall, cfg-word-overlap, semaphore-handshake, mailbox-sync, dataflow-cb-sync, srcreg-bank-sync, noc-sync, instruction-latency) across four synchronization surfaces, and add a cross-class JOIN pass that catches emergent races no single audit can see — where one audit's verdict is "safe because <invariant owned by another audit>". Use for a full hazard sweep of an LLK change, or before merging anything touching config writes, reconfig/uninit, inter-thread/cross-core sync, the SrcA/SrcB-Dst data path, or hand-written instruction sequences.
    125 repo stars
  76. ▌
    Mmio Race Audit · tenstorrent
    Audit LLK code for races between a RISC MMIO write to a config/GPR register and a Tensix instruction/MOP/replay that consumes it. Use after touching any raw cfg[...]=/reg_write/cfg_rmw/regfile[]= write near an UNPACR/PACR/MOP/CFGSHIFTMASK, or when adding addressing/stride/format register writes.
    125 repo stars
  77. ▌
    Test Driven Development · tenstorrent bundle
    Guides test-first workflows for bugs and features: write a minimal failing test that encodes contracts and clear failure messages, confirm the gap is not already covered, implement and document the fix, then verify green. Use when fixing bugs with tests, adding features via TDD, writing regression tests, or when the user asks for red-green-refactor or test-driven development.
    125 repo stars
  78. ▌
    Quasar Perf Test · tenstorrent
    Create, extend, debug, and validate Quasar LLK performance tests, their PerfRunType kernel paths, and tile/dimension sweep coverage. Use when adding perf_[op]_quasar.py, wiring PerfConfig, implementing UNPACK_ISOLATE / MATH_ISOLATE / PACK_ISOLATE / L1_CONGESTION, choosing PERF_TILE_SIZES or dest-fill matrices, or investigating implausible Quasar perf metrics and dvalid handshakes.
    125 repo stars
  79. ▌
    Mailbox Sync Audit · tenstorrent
    Audit LLK/compute-API use of the RISC↔RISC hardware mailboxes (mailbox_write/mailbox_read/mailbox_not_empty, TENSIX_MAILBOX*) for races/deadlock — push/pop balance per directed channel, call-count symmetry across threads, correct write-dest/read-src addressing, FIFO overflow, and the fence ordering caveat. Use after touching any mailbox_write/mailbox_read, the CB tile-address/value broadcast (circular_buffer.h / cb_api.h get_tile_address/read_tile_value), unpack-to-dest dst_index passing, or the debug halt/unhalt handshake.
    125 repo stars
  80. ▌
    Reconfig Stall Audit · tenstorrent
    Audit LLK reconfig/uninit/config-write functions for a MISSING stall that drains the execution unit before its config registers are rewritten (packer→PACK, unpacker→UNPACK, math→MATH|WAIT_SFPU). Use after touching cpack/cunpack/cmath, *_reconfig_*, *_uninit, set_packer_strides, or any function that writes ALU/THCON/ADDR_MOD/stride config.
    125 repo stars
  81. ▌
    Perf Parameter Impact · tenstorrent
    Analyze LLK performance-report CSV files and create evidence-backed parameter-impact reports in a Cursor Canvas. Use when asked how formats, fidelity, dimensions, accumulation, indexing, loop factors, or other parameters affect TILE_LOOP or another marker in a .post.csv report.
    125 repo stars
  82. ▌
    Perf Regression Check · tenstorrent bundle
    Compare one LLK perf test between two commits on this machine — by default the branch HEAD against the exact main commit it was branched from (the merge-base, not latest main), or any two refs the user names (hash, tag, origin/main, HEAD~5). Runs the sweep on both, several iterations each, and reports per-config regressions and improvements. Use when a developer asks "did my branch make perf worse", "check my changes for regressions", "compare my branch to where I branched off", or "compare perf between these two commits". Needs Tenstorrent hardware (runs the real perf sweep).
    125 repo stars
  83. ▌
    Cfg Word Overlap Audit · tenstorrent
    Audit LLK code for races on the backend CONFIG register file where differently-named fields share the SAME 32-bit config word — both cross-thread (unpack/math/pack write the same word) and intra-thread (a full-word write clobbers a sibling field the same thread set elsewhere). Use after adding/changing any ALU_FORMAT_SPEC / ALU_ACC_CTRL / ALU_ROUNDING_MODE / STACC_RELU / THCON_SEC* write, any WRCFG_32b/cfg[]= full-word write to a multi-field word, or any cfg_reg_rmw_tensix on a word another thread also touches.
    125 repo stars
  84. ▌
    Dataflow Cb Sync Audit · tenstorrent
    Audit circular-buffer (CB) producer/consumer flow control between data-movement (reader/writer) and compute kernels — cb_reserve_back/cb_push_back/cb_wait_front/cb_pop_front credit balance, data-write-before-credit ordering (NOC flush), reserve/wait-before-access, capacity vs num_pages, single-producer/consumer, counter cache-coherency, and remote/sharded CB credits. Use after touching any cb_* call, a reader/writer/compute kernel, fifo_rd_ptr/fifo_wr_ptr/pages_received/pages_acked, or RemoteSender/ReceiverCBInterface. Scope reaches beyond tt-llk into tt_metal/hw/inc/api/dataflow and ttnn/models kernels.
    125 repo stars
  85. ▌
    Srcreg Bank Sync Audit · tenstorrent
    Audit the shared backend DATA registers — SrcA/SrcB bank-valid (AllowedClient) + bank-flip handshake between unpacker and Matrix Unit, and the shared-once Dst/LReg overwrite hazards not already carried by the MATH_PACK semaphore or mutex::SFPU. Use after touching unpack→math dataflow, SETDVALID/CLEARDVALID, bank-flip bookkeeping, MOVD2A/MOVA2D/MOVB2D, or any cross-thread Dst/LReg access.
    125 repo stars
  86. ▌
    Perf Optimization Audit · tenstorrent bundle
    Audit Tensix/SFPU LLK compute kernels for PERFORMANCE — unfilled latency shadows/bubbles and redundant NOPs, redundant Dst/LReg store-load traffic, loop-invariant work, predication that should be branchless arithmetic (min/max/abs/setsgn), un-fused mul+add, ignored APPROXIMATION_MODE, and unroll/register-pressure mistakes. Use after touching any ckernel_sfpu_*.h, hand-written TTI_SFP*/TTI_* sequence, or the compute inner loop. This is a PERF audit (wasted cycles), NOT a correctness/race audit — pair it with instruction-latency-audit.
    125 repo stars
  87. ▌
    Instruction Latency Audit · tenstorrent
    Audit hand-written Tensix/SFPU instruction sequences for missing pipeline-latency padding — where a dependent instruction consumes a multi-cycle-latency result before it is ready and a NOP (or independent-instruction spacing) is required. Use after touching any raw TTI_SFP*/TTI_* sequence, ckernel_sfpu_* kernels, or hand-assembled instruction streams. NOT a cross-thread race — an intra-thread micro-architectural hazard.
    125 repo stars
  88. ▌
    Semaphore Handshake Audit · tenstorrent
    Audit LLK inter-thread synchronization (Tensix semaphores + ATGETM/ATRELM mutexes) for races/deadlock — SEMINIT correctness vs usage, post/get balance, wait-direction, RISC-MMIO-vs-Tensix ordering, and mutex acquire/release balance. Use after touching any t6_semaphore_*/semaphore_post/semaphore_get/SEMINIT/SEMWAIT/SEMPOST/SEMGET/t6_mutex_* or any math↔pack / unpack↔math handshake (MATH_PACK, UNPACK_TO_DEST, UNPACK_SYNC, MATH_DONE, FPU_SFPU).
    125 repo stars