Skip to content

first addition of skills#324

Open
AtlantaPepsi wants to merge 1 commit into
ROCm:developfrom
AtlantaPepsi:skill
Open

first addition of skills#324
AtlantaPepsi wants to merge 1 commit into
ROCm:developfrom
AtlantaPepsi:skill

Conversation

@AtlantaPepsi

Copy link
Copy Markdown
Contributor

Motivation

Technical Details

Test Plan

Test Result

Submission Checklist

Copilot AI left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Pull request overview

This PR introduces initial .claude/skills documentation and example artifacts to guide running and debugging TransferBench (including presets, config syntax, env vars, and multi-rank workflows).

Changes:

  • Added a transferbench-run skill with runnable recipes plus references for presets, env vars, and config-file grammar.
  • Added a transferbench-debug skill with symptom-driven troubleshooting guides and introspection helpers.
  • Added example configs and shell scripts for multi-node launch and topology/preset probing.

Reviewed changes

Copilot reviewed 12 out of 12 changed files in this pull request and generated 3 comments.

Show a summary per file
File Description
.claude/skills/transferbench-run/SKILL.md Run-side skill overview, invocation, recipes, and references.
.claude/skills/transferbench-run/references/presets.md Curated list of common presets and selection guidance.
.claude/skills/transferbench-run/references/env-vars.md Curated env-var guide organized by topic.
.claude/skills/transferbench-run/references/config-format.md Config grammar summary with examples and idioms.
.claude/skills/transferbench-run/examples/multi-node.sh Template mpirun launcher with env propagation.
.claude/skills/transferbench-run/examples/basic-p2p.cfg Minimal peer-to-peer config examples.
.claude/skills/transferbench-run/examples/advanced-mixed.cfg Advanced-mode config examples (per-transfer bytes/SEs).
.claude/skills/transferbench-debug/SKILL.md Debug-side skill overview and triage flow.
.claude/skills/transferbench-debug/references/verbose-introspection.md Debug-focused env vars + introspection command reference.
.claude/skills/transferbench-debug/references/multi-rank-debug.md Layered multi-rank (MPI/socket/RDMA) troubleshooting guide.
.claude/skills/transferbench-debug/references/common-failures.md Symptom → cause → fix catalog for common failures.
.claude/skills/transferbench-debug/examples/topology-probe.sh Script to capture banner/presets/envvars/help and optionally dump a preset.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Comment on lines +15 to +17
BINARY="${BINARY:-./TransferBench}"
[[ -x "./TransferBenchCuda" ]] && BINARY="${BINARY/TransferBench/TransferBenchCuda}"
[[ -x "$BINARY" ]] || { echo "ERROR: $BINARY not found or not executable"; exit 1; }
Comment on lines +10 to +12
# 1. Same binary on every node
for h in node0 node1; do ssh $h md5sum /home/timhu102/tBench/TransferBenchCuda; done

Comment on lines +1 to +4
---
name: transferbench-run
description: Use when the user wants to *run* TransferBench (the ROCm/CUDA memory-transfer benchmarking tool from AMD) — benchmarking, profiling, or measuring GPU/CPU/NIC bandwidth and latency. Covers writing config files, picking the right preset (a2a, p2p, sweep, nicp2p, podp2p, etc.), tuning environment variables, and launching single-node or multi-node (MPI / socket) runs. Does NOT cover building the binary from source, modifying its source code, or extending it with new presets/executors — for those, defer to a separate skill or the codebase itself.
---
@AtlantaPepsi AtlantaPepsi marked this pull request as ready for review June 18, 2026 21:56
@AtlantaPepsi AtlantaPepsi requested review from a team as code owners June 18, 2026 21:56
@AtlantaPepsi AtlantaPepsi changed the base branch from candidate-1.68 to develop June 18, 2026 21:56
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants