request-refactor-plan
通过用户访谈创建一个详细的重构计划,并将其拆分成很小的提交(tiny commits),最后以 GitHub issue 的形式归档。适用于用户希望规划一次重构、创建重构 RFC,或把重构拆成安全的渐进步骤。
Browse reusable Agent Skills, each with a clear purpose and practical guidance.
通过用户访谈创建一个详细的重构计划,并将其拆分成很小的提交(tiny commits),最后以 GitHub issue 的形式归档。适用于用户希望规划一次重构、创建重构 RFC,或把重构拆成安全的渐进步骤。
Cross-session handoff skill. Use it when the user explicitly wants to pause, switch sessions, resume prior work, restore a handoff package, or list recent handoffs.
Delivery-context organizer. Compress post-review changes into reviewer-ready PR and merge context without replacing release decisions.
Direction convergence skill. Use brainstorm, align, or challenge mode to turn messy questions or existing proposals into a clear directional decision.
Use this when direction is already clear and the next job is to compress it into an executable block with scope, done criteria, and routing.
Turn a raw feature spec into an approved Tasks Plan — grill the spec, have the Product Architect groom draft tasks (+ optional ADR and glossary additions), then run ONE human gate that decides everything touching the repo: tasks + storage, ADR, glossary, worktree, and which build to run.
Plan a refactor as an ordered, commit-grain Tasks Plan with structural acceptance criteria (suite green at every step, no behaviour diff) that `/squid-implement-night` can execute end-to-end.
Track daily PRs and Issues from vllm-project/vllm and vllm-project/vllm-ascend, filter by model (DeepSeek/Qwen/GLM/MiniMax/Kimi) and tech topics (PD disaggregation, MTP, quantization, graph mode, performance), analyze with LLM, and generate a Markdown report. Use when user wants vllm daily tracker, PR/Issue digest, or Ascend inference ecosystem monitoring.