AgentSkillsCN

sub-agent-delegation

将复杂任务委托给子代理，以实现并行自主工作。在GPU内核优化、数值正确性验证、性能剖析或长时间运行的验证需要专注独立执行时使用。

SKILL.md
--- frontmatter
name: sub-agent-delegation
description: Delegate complex tasks to sub-agents for parallel autonomous work. Use when GPU kernel optimization, numerical correctness verification, performance profiling, or long-running validation would benefit from focused independent execution.

Sub-Agent Delegation

Permissions

  • •NEVER spawn without explicit permission
  • •ASK first: "I've identified [TASK] for sub-agent delegation. Should I spawn one?"
  • •Explain WHY before requesting

When to Delegate

  • •GPU kernel optimization with iterative benchmarking
  • •Numerical correctness verification across test cases
  • •Performance profiling and analysis
  • •Parallel investigation of independent code paths
  • •Long-running validation suites

Patterns

  • •Parallel: Optimize independent kernels simultaneously (attention to A, MLP to B)
  • •Correctness First: Make tests pass before performance
  • •Incremental: Iterate until target speedup or report blockers

Kernel Optimization Template

code
Optimize [OPERATION] in [FILE].
Context: [current impl], [bottleneck source], [target HW: 3090/H100], [use case: train/inference]
Requirements:
1. Implement with Triton/CUDA
2. Verify: torch.allclose(atol=1e-5, rtol=1e-5), gradients match autograd
3. Benchmark: warmup=10, bench=100, report min/max/mean/std us
4. Scales: (1,128), (8,512), (32,2048)
Report: correctness status, perf table (scale, baseline_us, opt_us, speedup), memory

Workflow

Setup -> Develop -> Verify -> Benchmark -> Report

Requirements

  • •Report measured numbers, never estimates
  • •Include methodology (warmup, iterations, sync)
  • •Flag regressions immediately