create-optimize-pattern
Create new optimization pattern cards or symptom cards for the helix optimize knowledge base. Use when adding a new pattern card under skills/triton/triton-npu-optimize-knowledge/references/patterns/, skills/torch/torch-npu-optimize-knowledge/references/patterns/, skills/triton/triton-npu-cann-ext-api-patterns/references/patterns/, or a new symptom card under skills/triton/triton-npu-optimize-knowledge/references/symptoms/.
- License 许可
- Apache-2.0
- 所属仓库
- Piagari/helix
- 最近更新
- 2026-07-16
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- create-optimize-pattern
- description
- Create new optimization pattern cards or symptom cards for the helix optimize knowledge base. Use when adding a new pattern card under skills/triton/triton-npu-optimize-knowledge/references/patterns/, skills/torch/torch-npu-optimize-knowledge/references/patterns/, skills/triton/triton-npu-cann-ext-api-patterns/references/patterns/, or a new symptom card under skills/triton/triton-npu-optimize-knowledge/references/symptoms/.
03 SKILL.md 正文
Create Optimize Pattern / Symptom Card
Use this skill when creating new optimization pattern cards or symptom cards. Do not use for editing existing cards — edit those directly.
Target Directories
Place new cards under the appropriate directory:
| Card Type | Skill | Directory |
|---|---|---|
| Generic optimize pattern | triton-npu-optimize-knowledge |
skills/triton/triton-npu-optimize-knowledge/references/patterns/ |
| Torch NPU optimize pattern | torch-npu-optimize-knowledge |
skills/torch/torch-npu-optimize-knowledge/references/patterns/ |
| CANN extension API pattern | triton-npu-cann-ext-api-patterns |
skills/triton/triton-npu-cann-ext-api-patterns/references/patterns/ |
| Generic optimize symptom | triton-npu-optimize-knowledge |
skills/triton/triton-npu-optimize-knowledge/references/symptoms/ |
Pattern Card Template
---
priority: high|normal # optional — omit to default to normal
---
# <Human Title>
## Summary
## Use When
## Avoid When
## Signals
### Code
### Profile
### IR
## Related Patterns
## What To Verify After Applying
Pattern Card Rules
## Summaryand## Use Whenare required. All other predefined sections are optional.## Summarydescribes WHAT the pattern is/does. Keep it to 1-2 sentences. Do not include signal-like language or usage instructions.## Use Whendescribes WHEN to apply the pattern — detection conditions. Must be orthogonal to## Summary.priority: high|normalin frontmatter; omit to default tonormal.priorityonly affects generated index rendering. The generatedpattern_index.mdmust include a## High Priority Patternssection that lists cards markedhigh.## Use When,## Avoid When, and signal subsections work best as bullet lists. Both- itemand1. itemlist styles are accepted by the generator.- Free-form sections are allowed for examples, background, or architecture notes — they are ignored by the index generator but kept in the authored card.
Symptom Card Template
# <Human Title>
## Summary
## Evidence To Confirm
## Candidate Pattern Directions
## Common Non-Matches
Symptom Card Rules
## Summary,## Evidence To Confirm, and## Candidate Pattern Directionsare required.## Common Non-Matchesis optional.- Keep predefined section names exact.
After Creation
- Regenerate the affected index files:
bash scripts/update-optimize-knowledge-indices.sh
- Never hand-edit
skills/*/references/pattern_index.mdorskills/*/references/symptom_index.md— they are generated.
Reference
For detailed authoring notes, see docs/notes/2026-04-29-optimize-pattern-card-authoring.md.
04 同仓库更多 Skills
triton-npu-repair-guide
Heuristic fixes for Ascend Triton compile/JIT/kernel errors and numerical precision mismatches when editing or converting the operator. Does not replace ascend-npu-gen-test, ascend-npu-gen-bench, or harness specs.
triton-npu-optimize
Iteratively optimize a Triton Ascend NPU operator with correctness and performance gates. Use for operator optimization tasks that need repeated correctness validation, benchmark validation, multi-round experiment tracking, reusable optimization notes, and profiler-backed performance analysis when benchmark results need deeper explanation.
triton-npu-optimize-knowledge
Generic reference-only optimize knowledge for pattern triage and evidence-backed symptom routing. This skill does not define optimize workflow or own round artifacts.
triton-npu-optimize-knowledge-v3
Generic reference-only optimize knowledge (v3 working copy) for pattern triage and evidence-backed symptom routing. This skill does not define optimize workflow or own round artifacts; prefer this tree for ongoing knowledge updates.
triton-npu-optimize-knowledge
Generic reference-only optimize knowledge for pattern triage and evidence-backed symptom routing. This skill does not define optimize workflow or own round artifacts.
triton-npu-convert-pytorch-operator
Convert one PyTorch operator into a Triton NPU-backed PyTorch operator, preserve the trailing input-helper block, and validate the converted output through standalone or differential testing.