dude1wudv

Project-local lean coding mode. Use for SRAD edits to choose the shortest correct change, avoid speculative abstractions, prefer existing code and standard tools, and keep checks minimal but real.

Project-local planning/docs skill for SRAD. Use when user input is vague, requirements are ambiguous, a task needs PRD/spec/design/tasks/docs, or the user says plan-docs/plan-docx/Kiro/spec/需求/规划/文档/开发追踪. Automatically turn fuzzy input into…

Alias for project-local plan-docs. Use when the user writes plan-docx, plan-docs, fuzzy requirements, PRD/spec/design/tasks, or asks to turn vague input into reusable SRAD workflow documentation.

Project-local skill creation/update guide. Use when adding or updating skills under .agents/skills so GitHub collaborators get the same SRAD workflow skills.

Project-local spec-driven workflow for SRAD. Use when a task mentions Kiro, specs, requirements/design/tasks, subagents, context-window overflow, large file reads, Get-Content, or spans 3+ files. Prefer durable .kiro/specs plus bounded agents or targeted…

Use when changing, optimizing, debugging, compiling, testing, or documenting variants/srad/ours_192lane in this SRAD VCK190 repository. Enforces project-local code map first, Kiro-style change docs for multi-step work, Superpowers red-green TDD before…

Project-local copy of the Superpowers plan execution skill. Use when executing a written SRAD implementation plan task-by-task with TDD and verification checkpoints.

Project-local Superpowers companion for executing SRAD plans with bounded subagents. Use when a written plan has independent tasks or when large-file reading should be delegated to scouts.

Use for CMake build files, configuration, generators, presets, and C/C++ build troubleshooting.

Use for `E:\See-you-more-than-her` or `E:\codex-governance` publish/report runs that mention zhongshu plans, existing PR updates, local launcher APIs, or "submit all local changes" without broad extra checking.

Git 工作流自动化 - 智能分支管理、提交信息生成、PR 创建

Default lean coding mode for most code edits, bug fixes, UI tweaks, refactors, and small feature work. Use automatically unless the user requests exhaustive tests, broad refactoring, deep audit/review, high-stakes security work, migration-heavy changes, or…

Use when complex work needs a SelfForge avatar to supervise Codex, review another agent, catch preference drift, or make preference-based decisions with retrieved local evidence.

Use for C/C++ static analysis with clang-tidy, cppcheck, scan-build, and warning triage.

自动生成单元测试：识别函数逻辑、覆盖边界条件、生成可运行测试代码

Use for webpage frontend design and implementation.

为项目生成可执行的波次计划文档。当用户进入 Plan 模式、要求开发计划、任务拆分、 并行执行、依赖排序、实施波次或 AI/Subagent 协作时使用。简单且范围明确的短任务 保持主线执行；跨模块、调用链未知、存在兼容性或多个独立工作面的任务生成依赖图、 Wave 顺序、并行边界和完成门。兼容 Cursor Agent Skills 与 cursor-byok-source。

本地模式实现指南

Use when extracting, formatting, refreshing, or investigating a read-only formatted snapshot of the installed Cursor.app bundle under .cursor-app-formatted; includes git-ignore rules, snapshot generation workflow, and the rule to inspect formatted code…

Use when debugging Cursor client agent/local-mode/tool/backend-store/provider-replay failures in this repo, especially after the state/context history-store refactor, when triaging installed app bundles read-only, correlating installed-client behavior with…

当需要调查 Cursor 本地模式 debug/log 证据时使用：config.yaml 的 log 热加载、history/<conversationId>/debug JSONL 文件、Bidi 原始/解码记录、RunSSE 记录、runtime/provider debug 记录、debug 文件缺失原因，或解释这些 debug 文件如何生成与如何查询。

Use when changing prompt compilation, history replay, persisted conversation state, model request construction, or dynamic reminders in this repo; protects prefix-cache hit rate by keeping model-visible history append-only and dynamic attention scoped to the…

Automatically schedule independent Sentaurus SDevice simulations on separate VM CPU cores with one thread per case. Use for every project SDevice run, parameter sweep, LET scan, temperature scan, bias scan, or concurrent TCAD batch. Keep mesh/restart…

Run Synopsys Sentaurus TCAD simulations on the local Linux VM over SSH, with project-level fail-closed automatic single-core leases for independent SDevice jobs. Use when the user asks Codex to connect to tcad@192.168.137.131, start or diagnose…

Run a dynamic-workflow script on a local Codex App Server — orchestrate many Codex / GPT agents (the agent / parallel / pipeline / phase / budget DSL) instead of Claude subagents, for codebase audits, large migrations, and multi-agent review or research. Give…

项目规划与文档化技能。当用户要为一个项目做需求梳理、产品规划、建立/完善文档体系、 把模糊想法细化成可开发的功能定义、或想约束 AI 按文档精确实现时使用。以产品经理视角 讨论需求与实用性，把需求拆解到「功能模块→子模块→按钮级交互」的递归文档树，并据此生成 高保真原型、开发追踪与可直接交给 AI 执行的微观化任务提示词。支持新项目从零引导， 也支持老项目逆向构建文档树。覆盖：项目规划、PRD、需求文档、架构文档、原型设计、 任务拆分、开发计划。Use when planning a project,…

Use when setting up or repairing host-to-VM control of Cadence Virtuoso through VMware, CentOS/Linux SSH, virtuoso-bridge-lite, Spectre, SKILL CIW load, or when errors mention NO-CARRIER, VMware NAT/DHCP stopped, daemon NO RESPONSE, RBStop in bash, or…