ExecuTorch Binary Size Reduction
pytorch/executorch
Measures and shrinks the ExecuTorch runtime binary by building a size test, analyzing it with bloaty and landing each reduction as its own pull request.
Spots and fixes known x86 C and C++ performance problems from source or profiler output, such as false sharing, missing restrict and narrow SIMD.
Install x86 C and C++ Performance Patterns
$ npx skills add intel/intel-performance-skills --skill performance-patterns -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install intel/intel-performance-skills performance-patterns --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/intel/intel-performance-skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/performance-patterns .claude/skills/performance-patterns && 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 "performance-patterns" agent skill from https://github.com/intel/intel-performance-skills/tree/main/skills/performance-patterns into .claude/skills/performance-patterns/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "performance-patterns", 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/intel/intel-performance-skills/tree/main/skills/performance-patternsType 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 intel/intel-performance-skills --skill performance-patterns -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install intel/intel-performance-skills performance-patterns --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/intel/intel-performance-skills.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/performance-patterns .agents/skills/performance-patterns && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "performance-patterns" agent skill from https://github.com/intel/intel-performance-skills/tree/main/skills/performance-patterns into .agents/skills/performance-patterns/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "performance-patterns", 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 intel/intel-performance-skills --skill performance-patterns -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install intel/intel-performance-skills performance-patterns --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/intel/intel-performance-skills.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/performance-patterns .cursor/skills/performance-patterns && 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 "performance-patterns" agent skill from https://github.com/intel/intel-performance-skills/tree/main/skills/performance-patterns into .cursor/skills/performance-patterns/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "performance-patterns", 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/intel/intel-performance-skills.git --path skills/performance-patterns--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 intel/intel-performance-skills --skill performance-patterns -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install intel/intel-performance-skills performance-patterns --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/intel/intel-performance-skills.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/performance-patterns .gemini/skills/performance-patterns && 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 "performance-patterns" agent skill from https://github.com/intel/intel-performance-skills/tree/main/skills/performance-patterns into .gemini/skills/performance-patterns/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "performance-patterns", 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 intel/intel-performance-skills performance-patternsInstalls 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 intel/intel-performance-skills --skill performance-patterns -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/intel/intel-performance-skills.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/performance-patterns .github/skills/performance-patterns && 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 "performance-patterns" agent skill from https://github.com/intel/intel-performance-skills/tree/main/skills/performance-patterns into .github/skills/performance-patterns/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "performance-patterns", 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 intel/intel-performance-skills --skill performance-patterns -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install intel/intel-performance-skills performance-patterns --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/intel/intel-performance-skills.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/performance-patterns .opencode/skills/performance-patterns && 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 "performance-patterns" agent skill from https://github.com/intel/intel-performance-skills/tree/main/skills/performance-patterns into .opencode/skills/performance-patterns/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "performance-patterns", 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
performance-patterns- GitHub stars
- 332
- Token cost
- ~903 tokens
- SKILL.md length
- 326 words
- Files
- 33 (incl. references)
- Skills in repo
- 3
- Repo updated
- First seen
- Licence
- Custom licence
At a glance
Spots and fixes known x86 C and C++ performance problems from source or profiler output, such as false sharing, missing restrict and narrow SIMD.
- Works in 4 steps: Load the right file for your context → Identify the matching pattern → Read the pattern detail file → …
- Reading a perf or VTune profile and finding which known pattern it shows
- SKILL.md covers How to use this skill and Reusable library modules
- Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
What it does
A catalog of well-known performance problems in x86, C and C++ code, each with detection signals and a step-by-step resolution. The skill routes by context: profiling output from perf, VTune or flamegraphs goes to triggers/from-profile.md, existing source with no profile goes to triggers/from-source.md, and writing new performance-sensitive or SIMD code goes to guidelines/new-code.md.
Patterns include serial accumulator loops, TTAS spinlocks, narrow SIMD that could be wider, false sharing, per-CPU statistics, missing vzeroupper or restrict, condition-variable thundering herd, mutex-to-rwlock changes, CPU dispatch, library version upgrades, fast CRC32C and known algorithms such as cosine similarity, Hamming and Jaccard distance, plus x86-simd-sort. When a pattern matches, the agent reads that pattern's file instead of fixing from memory, applies the fix and runs its verification method.
Several patterns can apply at once, so all plausible matches are checked first. Reusable library modules cover runtime CPU dispatch with target_clones or __builtin_cpu_supports, a zipper algorithm for widening vector registers, and CRC32C implementations in portable, SSE and AVX-512 variants.
When your agent uses it
- Reading a perf or VTune profile and finding which known pattern it shows
- Reviewing C or C++ source for false sharing or serial reduction loops
- Writing a fast dot product, reduction or CRC32C routine
- Adding runtime CPU dispatch to a function with several vectorized variants
Example prompts
- “This perf annotate output shows a hot cmpxchg cluster. Which performance pattern fits?”
- “Review this C++ reduction loop for performance problems before we ship it.”
- “Write a CPU-dispatched dot product with an AVX-512 path and a portable fallback.”
- “Our worker threads wake all waiters on every notify. Is this the thundering herd pattern?”
Requirements
- C or C++ source, or profiler output from perf, VTune or flamegraphs
- A compiler and benchmark setup to verify fixes
Workflow steps
4 steps, taken from the step headings in SKILL.md.
- Load the right file for your context
- Identify the matching pattern
- Read the pattern detail file
- Apply the fix and verify
What it can do on your machine
Read from SKILL.md and the folder at commit 3e33aec. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
No scripts in the folder and no shell commands in SKILL.md.
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
x86 C and C++ Performance Patterns loads about 903 tokens when it runs, and up to ~3.8k if it reads all its reference files. Until then it costs about 236 tokens; SKILL.md has 326 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 passed
The automated check found no risky patterns in SKILL.md.
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 326 words (~903 tokens).
“A growing catalog of well-known code patterns that cause performance problems, with detection signals and resolution playbooks for each.”
- name
- performance-patterns
Files
SKILL.md and 32 other files (references) in skills/performance-patterns of intel/intel-performance-skills.
- SKILL.md
- design.md
- guidelines/new-code.md
- library/cpu-dispatch.md
- library/crc32c_avx512_vpclmulqdq.c
- library/crc32c_portable.c
- library/crc32c_sse42.c
- patterns/cold-path-annotation.md
- patterns/cv-thundering-herd.md
- patterns/false-sharing.md
- patterns/fast-crc32c-impl.md
- patterns/fast-crc32c.md
- patterns/library-version-upgrade.md
- patterns/missing-restrict.md
- patterns/missing-vzeroupper.md
- patterns/mutex-to-rwlock.md
- patterns/parallel-accumulator.md
- patterns/per-cpu-stats.md
- … and 15 more
Open the folder on GitHubat commit 3e33aec
Compare with similar skills
x86 C and C++ Performance Patterns 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 |
|---|---|---|---|---|---|---|
| x86 C and C++ Performance Patterns this skillintel/intel-performance-skills | 332 | — | ~903 | Automated safety check: Pass | Custom licence | |
| ExecuTorch Binary Size Reductionpytorch/executorch | 5.1k | — | ~793 | Automated safety check: Pass | Custom licence | |
| Cppcrazyguitar/cppcheatsheet | 290 | — | ~1.8k | Automated safety check: Pass | MIT | |
| Ripwire Optimization Remarks Triageredhat-et/ripwire | 2.4k | — | ~3.1k | Automated safety check: Notes | Apache-2.0 | |
| Cpp Prodavila7/claude-code-templates | 32k | 7 repos | ~467 | Automated safety check: Pass | MIT | |
| Interval Profiling Performance AnalyzerArabelaTso/Skills-4-SE | 253 | — | ~1.7k | Automated safety check: Notes | Apache-2.0 |
Similar skills
Cpp
crazyguitar/cppcheatsheet
Comprehensive C/C++ programming reference covering everything from C11-C23 and C++11-C++23, system programming, CUDA GPU computing, debugging tools, Rust interop, and advanced topics.
Ripwire Optimization Remarks Triage
redhat-et/ripwire
Contributor guide for reading clang optimization remarks while editing ripwire's own C++, deciding between a source change and a build change such as LTO or PGO.
Cpp Pro
davila7/claude-code-templates
Write idiomatic C++ code with modern features, RAII, smart pointers, and STL algorithms.
Interval Profiling Performance Analyzer
ArabelaTso/Skills-4-SE
Profile programs at the function/method level to identify performance hotspots, bottlenecks, and optimization opportunities.
PlotJuggler Live Verification
PlotJuggler/PlotJuggler
Confirms that a change to PlotJuggler 4 really works in the running app by proving the rebuild, launching with real data and measuring the result.
More from intel/intel-performance-skills
- Official
Linux perf Profiling Guide
intel/intel-performance-skills
Guides perf-based profiling of Linux programs through five flows: quick hardware counters, hotspot profiling, cache contention, core-scaling, and a report.
- Official
Phoronix Test Suite Benchmarks
intel/intel-performance-skills
Installs, runs, parses and optimizes Phoronix Test Suite benchmarks, saving scores and comparing results before and after code changes.
Related
Works with
Categories
Questions about x86 C and C++ Performance Patterns
What does x86 C and C++ Performance Patterns do?
Spots and fixes known x86 C and C++ performance problems from source or profiler output, such as false sharing, missing restrict and narrow SIMD. A catalog of well-known performance problems in x86, C and C++ code, each with detection signals and a step-by-step resolution.md.
When should I use x86 C and C++ Performance Patterns?
x86 C and C++ Performance Patterns fits situations like: reading a perf or VTune profile and finding which known pattern it shows; reviewing C or C++ source for false sharing or serial reduction loops; writing a fast dot product, reduction or CRC32C routine; adding runtime CPU dispatch to a function with several vectorized variants.
How do I install x86 C and C++ Performance Patterns in Claude Code?
Run `npx skills add intel/intel-performance-skills --skill performance-patterns -a claude-code`. Or copy the skill folder (skills/performance-patterns in intel/intel-performance-skills) into .claude/skills/performance-patterns in your project. Claude Code loads it when a task matches its description.
How do I install x86 C and C++ Performance Patterns in Codex?
Run `npx skills add intel/intel-performance-skills --skill performance-patterns -a codex`. Or copy the skill folder (skills/performance-patterns in intel/intel-performance-skills) into .agents/skills/performance-patterns in your project. Codex loads it when a task matches its description.
Can I use x86 C and C++ Performance Patterns 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 intel/intel-performance-skills --skill performance-patterns -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/performance-patterns, .gemini/skills/performance-patterns, .github/skills/performance-patterns and .opencode/skills/performance-patterns in your project.
What does x86 C and C++ Performance Patterns need to run?
SKILL.md names no scripts, command-line tools or credentials: x86 C and C++ Performance Patterns is instructions for the agent only. Our summary lists: C or C++ source, or profiler output from perf, VTune or flamegraphs; A compiler and benchmark setup to verify fixes.
Does x86 C and C++ Performance Patterns 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 x86 C and C++ Performance Patterns safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does x86 C and C++ Performance Patterns use?
x86 C and C++ Performance Patterns 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 x86 C and C++ Performance Patterns use?
About 903 tokens (SKILL.md is roughly 3.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 2.9k tokens, read only when the agent opens those files.
What are the alternatives to x86 C and C++ Performance Patterns?
Skills that share tags, products or a category with x86 C and C++ Performance Patterns: ExecuTorch Binary Size Reduction (pytorch/executorch, 5.1k stars), Cpp (crazyguitar/cppcheatsheet, 290 stars), Ripwire Optimization Remarks Triage (redhat-et/ripwire, 2.4k stars) and Cpp Pro (davila7/claude-code-templates, 32k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains x86 C and C++ Performance Patterns?
intel (a GitHub organization, an official publisher) maintains it in intel/intel-performance-skills, which has 332 GitHub stars. The repository holds 3 skills in this directory. The repository was last updated on September 28, 2026.
Source: intel/intel-performance-skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.