Megatron-LM on SLURM
NVIDIA/Megatron-LM
Shows how to launch distributed Megatron-LM training on a SLURM cluster: sbatch skeleton, torch.distributed.run setup, CUDA_DEVICE_MAX_CONNECTIONS rules and failure diagnosis.
Detect out-of-bounds memory accesses in CPU or GPU code using static interval analysis and runtime assertions/printfs.
Install Oob Detection
$ npx skills add ROCm/FlyDSL --skill oob-detection -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install ROCm/FlyDSL oob-detection --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/ROCm/FlyDSL.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/oob-detection .claude/skills/oob-detection && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "oob-detection" agent skill from https://github.com/ROCm/FlyDSL/tree/main/.claude/skills/oob-detection into .claude/skills/oob-detection/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "oob-detection", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/ROCm/FlyDSL/tree/main/.claude/skills/oob-detectionType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add ROCm/FlyDSL --skill oob-detection -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install ROCm/FlyDSL oob-detection --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/ROCm/FlyDSL.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.claude/skills/oob-detection .agents/skills/oob-detection && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "oob-detection" agent skill from https://github.com/ROCm/FlyDSL/tree/main/.claude/skills/oob-detection into .agents/skills/oob-detection/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "oob-detection", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add ROCm/FlyDSL --skill oob-detection -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install ROCm/FlyDSL oob-detection --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/ROCm/FlyDSL.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.claude/skills/oob-detection .cursor/skills/oob-detection && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "oob-detection" agent skill from https://github.com/ROCm/FlyDSL/tree/main/.claude/skills/oob-detection into .cursor/skills/oob-detection/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "oob-detection", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/ROCm/FlyDSL.git --path .claude/skills/oob-detection--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add ROCm/FlyDSL --skill oob-detection -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install ROCm/FlyDSL oob-detection --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/ROCm/FlyDSL.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.claude/skills/oob-detection .gemini/skills/oob-detection && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "oob-detection" agent skill from https://github.com/ROCm/FlyDSL/tree/main/.claude/skills/oob-detection into .gemini/skills/oob-detection/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "oob-detection", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install ROCm/FlyDSL oob-detectionInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add ROCm/FlyDSL --skill oob-detection -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/ROCm/FlyDSL.git skills-src && mkdir -p .github/skills && cp -r skills-src/.claude/skills/oob-detection .github/skills/oob-detection && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "oob-detection" agent skill from https://github.com/ROCm/FlyDSL/tree/main/.claude/skills/oob-detection into .github/skills/oob-detection/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "oob-detection", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add ROCm/FlyDSL --skill oob-detection -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install ROCm/FlyDSL oob-detection --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/ROCm/FlyDSL.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.claude/skills/oob-detection .opencode/skills/oob-detection && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "oob-detection" agent skill from https://github.com/ROCm/FlyDSL/tree/main/.claude/skills/oob-detection into .opencode/skills/oob-detection/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "oob-detection", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
oob-detection- GitHub stars
- 290
- Token cost
- ~1k tokens
- SKILL.md length
- 419 words
- Files
- 1
- Skills in repo
- 19
- Repo updated
- First seen
- Licence
- Custom licence
At a glance
Detect out-of-bounds memory accesses in CPU or GPU code using static interval analysis and runtime assertions/printfs.
- Works in 5 steps: Classify the OOB → Static Interval Analysis → Add Runtime Logical Checks → …
- Investigating OOB
- SKILL.md covers 1. Classify the OOB, 2. Static Interval Analysis, 3. Add Runtime Logical Checks and 4. Fix Strategy, plus 1 more section
- Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
What it does
Oob Detection is an agent skill from ROCm/FlyDSL. Detect out-of-bounds memory accesses in CPU or GPU code using static interval analysis and runtime assertions/printfs. Use when investigating OOB, buffer overrun, invalid memory access, HIP/ROCm illegal address, CUDA illegal memory access, silent tensor corruption, or suspicious bufferload/store address arithmetic.
Its SKILL.md is about 1k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.
It sits in Security. It works with CUDA. The repository describes itself as: FlyDSL is the Python front‑end of the project: a Flexible Layout Python DSL for expressing tiling, partitioning, data movement, and kernel structure at a high level.
When your agent uses it
- Investigating OOB
- Invalid memory access
- HIP/ROCm illegal address
- CUDA illegal memory access
Example prompts
- “/oob-detection”
Requirements
- Python 3
- Pre-approved tools (allowed-tools): Read, Edit, Bash, Grep, Glob, Agent
Workflow steps
5 steps, taken from the step headings in SKILL.md.
- Classify the OOB
- Static Interval Analysis
- Add Runtime Logical Checks
- Fix Strategy
- Report Format
What it can do on your machine
Read from SKILL.md and the folder at commit 3ad47c1. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves these tools, so the agent can use them without asking each time:
ReadEditBashGrepGlobAgent
From allowed-tools in the SKILL.md frontmatter.
Runs code
No scripts in the folder and no shell commands in SKILL.md (its code samples are python).
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Oob Detection loads about 1k tokens when it runs. Until then it costs about 83 tokens; SKILL.md has 419 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check: notes
The automated check noted patterns worth knowing about, such as sudo or a known installer.
- NotePre-approves every shell command (allowed-tools: Bash)SKILL.md
allowed-tools: Read, Edit, Bash, Grep, Glob, Agent
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
Its licence (Custom licence) doesn't allow us to republish the file, so here is its outline and opening line. It has 419 words (~1,031 tokens).
“Use this skill when a kernel, runtime, or host path may read or write outside its intended buffer or logical tile. Prefer proving the address range first instead of relying only on runtime failures.”
- name
- oob-detection
- allowed-tools
- Read, Edit, Bash, Grep, Glob, Agent
Files
Just SKILL.md in .claude/skills/oob-detection of ROCm/FlyDSL.
Open the folder on GitHubat commit 3ad47c1
Compare with similar skills
Oob Detection next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Oob Detection this skillROCm/FlyDSL | 290 | — | ~1k | Automated safety check: Notes | Custom licence | |
| Megatron-LM on SLURMNVIDIA/Megatron-LM | 18k | — | ~1.8k | Automated safety check: Pass | Apache-2.0 | |
| Cosmos3 Env TroubleshootNVIDIA/cosmos-framework | 559 | — | ~1.3k | Automated safety check: Notes | Custom licence | |
| Install Miles Diffusionradixark/miles_diffusion | 107 | — | ~1.6k | Automated safety check: Pass | Apache-2.0 | |
| Init GPU Serverdrawthingsai/draw-things-community | 582 | — | ~2.2k | Automated safety check: Pass | GPL-3.0 | |
| Cosmos3 SetupNVIDIA/cosmos-framework | 559 | — | ~1.1k | Automated safety check: Notes | Custom licence |
Similar skills
- Official
- Official
Cosmos3 Env Troubleshoot
NVIDIA/cosmos-framework
Diagnose and fix Cosmos3 environment, installation, and runtime errors.
Install Miles Diffusion
radixark/miles_diffusion
Fallback installer for milesdiffusion on a bare CUDA 12.9 Linux GPU box, reproducing the official radixark/milesdiffusion image's package versions and verifying them.
Init GPU Server
drawthingsai/draw-things-community
Initialize a Draw Things GPU server with GPUScript, including script sync, Docker/CUDA/NVIDIA runtime setup, 7T data disk mounting, mergerfs, and end-to-end GPU verification.
- Official
Cosmos3 Setup
NVIDIA/cosmos-framework
Guide users through Cosmos3 installation, environment setup, checkpoint downloading, and verification.
- Official
Kermt Setup
NVIDIA/skills
Bootstrap the KERMT agent environment — verify host docker + nvidia-container-toolkit, build the kermt:latest image from the repo's Dockerfile if it doesn't yet exist, and run a GPU smoke test…
More from ROCm/FlyDSL
All 19 skills in this repoLlvm
ROCm/FlyDSL
Tune and analyse a FlyDSL kernel at the LLVM level: pick a compile hint, function attribute, or AMDGPU backend flag, then PROVE it reached codegen.
Isa Resource Diff
ROCm/FlyDSL
Detect per-kernel GPU resource regressions (VGPR, SGPR, register spills, scratch, static LDS) by diffing the final ISA before and after a change, using its isaresourcetable.py helper.
API Stability
ROCm/FlyDSL
Review a FlyDSL PR, commit, branch, kernel, or consuming module for API-stability compliance.
Build Rocm Image
ROCm/FlyDSL
Connect to a remote host via SSH and build a Docker image with rocprofv3, aiter, and FlyDSL.
Debug Flydsl Kernel
ROCm/FlyDSL
Debug FlyDSL GPU kernels that produce NaN, inf, wrong results, or crash.
Flydsl Tile Programming
ROCm/FlyDSL
Guided step-by-step wizard for producing a new FlyDSL GPU kernel from a requirement: classify the kernel type, pick a skeleton, fill in compute, add control flow / sync / LDS, then test on GPU.
Related
Works with
Categories
Questions about Oob Detection
What does Oob Detection do?
Detect out-of-bounds memory accesses in CPU or GPU code using static interval analysis and runtime assertions/printfs. Oob Detection is an agent skill from ROCm/FlyDSL. Detect out-of-bounds memory accesses in CPU or GPU code using static interval analysis and runtime assertions/printfs.
When should I use Oob Detection?
Oob Detection fits situations like: investigating OOB; invalid memory access; HIP/ROCm illegal address; CUDA illegal memory access.
How do I install Oob Detection in Claude Code?
Run `npx skills add ROCm/FlyDSL --skill oob-detection -a claude-code`. Or copy the skill folder (.claude/skills/oob-detection in ROCm/FlyDSL) into .claude/skills/oob-detection in your project. Claude Code loads it when a task matches its description.
How do I install Oob Detection in Codex?
Run `npx skills add ROCm/FlyDSL --skill oob-detection -a codex`. Or copy the skill folder (.claude/skills/oob-detection in ROCm/FlyDSL) into .agents/skills/oob-detection in your project. Codex loads it when a task matches its description.
Can I use Oob Detection in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add ROCm/FlyDSL --skill oob-detection -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/oob-detection, .gemini/skills/oob-detection, .github/skills/oob-detection and .opencode/skills/oob-detection in your project.
What does Oob Detection need to run?
SKILL.md names no scripts, command-line tools or credentials: Oob Detection is instructions for the agent only. Our summary lists: Python 3. Its frontmatter pre-approves these tools: Read, Edit, Bash, Grep, Glob, Agent.
Does Oob Detection access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Oob Detection safe to install?
Our automated static check of SKILL.md found notes only (pre-approves every shell command (allowed-tools: bash)), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.
What licence does Oob Detection use?
Oob Detection has a licence file (the repository's licence) that doesn't match a standard licence. Read it on GitHub before reusing the skill.
How many tokens does Oob Detection use?
About 1k tokens (SKILL.md is roughly 4.1k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Oob Detection?
Skills that share tags, products or a category with Oob Detection: Megatron-LM on SLURM (NVIDIA/Megatron-LM, 18k stars), Cosmos3 Env Troubleshoot (NVIDIA/cosmos-framework, 559 stars), Install Miles Diffusion (radixark/miles_diffusion, 107 stars) and Init GPU Server (drawthingsai/draw-things-community, 582 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Oob Detection?
ROCm (a GitHub organization) maintains it in ROCm/FlyDSL, which has 290 GitHub stars. The repository holds 19 skills in this directory. The repository was last updated on October 9, 2026.
Source: ROCm/FlyDSL on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.