User Story Writer
deanpeters/Product-Manager-Skills
Writes user stories in Mike Cohn's format with Gherkin acceptance criteria, turning user needs into development-ready work with testable conditions.
让 Codex 与 Claude Code 通过本机 CLI 组成“主执行者 + 独立审查者”的对称协作闭环。用于用户要求两个模型一起完成、交叉校核、独立审查、修到 ACK,或任务涉及产品需求收敛、复杂方案、跨模块开发、迁移、安全、重要重构和高质量交付时;无论从 Codex 还是 Claude Code 启动,都由当前模型主持,并完整调用另一方完成需求挑战、方案门禁、实现冷审和最终验收。
$ npx skills add yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install yunshu0909/yunshu_skillshub dual-agent-collaboration --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/yunshu0909/yunshu_skillshub.git skills-src && mkdir -p .claude/skills && cp -r skills-src/coding/dual-agent-collaboration .claude/skills/dual-agent-collaboration && 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 "dual-agent-collaboration" agent skill from https://github.com/yunshu0909/yunshu_skillshub/tree/master/coding/dual-agent-collaboration into .claude/skills/dual-agent-collaboration/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "dual-agent-collaboration", 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/yunshu0909/yunshu_skillshub/tree/master/coding/dual-agent-collaborationType 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 yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install yunshu0909/yunshu_skillshub dual-agent-collaboration --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/yunshu0909/yunshu_skillshub.git skills-src && mkdir -p .agents/skills && cp -r skills-src/coding/dual-agent-collaboration .agents/skills/dual-agent-collaboration && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "dual-agent-collaboration" agent skill from https://github.com/yunshu0909/yunshu_skillshub/tree/master/coding/dual-agent-collaboration into .agents/skills/dual-agent-collaboration/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "dual-agent-collaboration", 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 yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install yunshu0909/yunshu_skillshub dual-agent-collaboration --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/yunshu0909/yunshu_skillshub.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/coding/dual-agent-collaboration .cursor/skills/dual-agent-collaboration && 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 "dual-agent-collaboration" agent skill from https://github.com/yunshu0909/yunshu_skillshub/tree/master/coding/dual-agent-collaboration into .cursor/skills/dual-agent-collaboration/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "dual-agent-collaboration", 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/yunshu0909/yunshu_skillshub.git --path coding/dual-agent-collaboration--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 yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install yunshu0909/yunshu_skillshub dual-agent-collaboration --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/yunshu0909/yunshu_skillshub.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/coding/dual-agent-collaboration .gemini/skills/dual-agent-collaboration && 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 "dual-agent-collaboration" agent skill from https://github.com/yunshu0909/yunshu_skillshub/tree/master/coding/dual-agent-collaboration into .gemini/skills/dual-agent-collaboration/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "dual-agent-collaboration", 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 yunshu0909/yunshu_skillshub dual-agent-collaborationInstalls 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 yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/yunshu0909/yunshu_skillshub.git skills-src && mkdir -p .github/skills && cp -r skills-src/coding/dual-agent-collaboration .github/skills/dual-agent-collaboration && 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 "dual-agent-collaboration" agent skill from https://github.com/yunshu0909/yunshu_skillshub/tree/master/coding/dual-agent-collaboration into .github/skills/dual-agent-collaboration/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "dual-agent-collaboration", 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 yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install yunshu0909/yunshu_skillshub dual-agent-collaboration --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/yunshu0909/yunshu_skillshub.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/coding/dual-agent-collaboration .opencode/skills/dual-agent-collaboration && 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 "dual-agent-collaboration" agent skill from https://github.com/yunshu0909/yunshu_skillshub/tree/master/coding/dual-agent-collaboration into .opencode/skills/dual-agent-collaboration/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "dual-agent-collaboration", 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.
dual-agent-collaboration让 Codex 与 Claude Code 通过本机 CLI 组成“主执行者 + 独立审查者”的对称协作闭环。用于用户要求两个模型一起完成、交叉校核、独立审查、修到 ACK,或任务涉及产品需求收敛、复杂方案、跨模块开发、迁移、安全、重要重构和高质量交付时;无论从 Codex 还是 Claude Code 启动,都由当前模型主持,并完整调用另一方完成需求挑战、方案门禁、实现冷审和最终验收。
Dual Agent Collaboration is an agent skill from yunshu0909/yunshu_skillshub. 让 Codex 与 Claude Code 通过本机 CLI 组成“主执行者 + 独立审查者”的对称协作闭环。用于用户要求两个模型一起完成、交叉校核、独立审查、修到 ACK,或任务涉及产品需求收敛、复杂方案、跨模块开发、迁移、安全、重要重构和高质量交付时;无论从 Codex 还是 Claude Code 启动,都由当前模型主持,并完整调用另一方完成需求挑战、方案门禁、实现冷审和最终验收。
Its SKILL.md is about 2.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 10 other files, including scripts and reference files (for example `agents/openai.yaml`, `references/handoff-template.md` and `references/review-output.schema.json`).
It sits in Product & Project Management. The repository describes itself as: 云舒精选的 Claude Code Skills 集合,提升开发和产品管理效率. The licence is MIT.
7 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit da0d31f. It shows what the files ask for, not the result of running them.
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.
Ships 4 files in scripts/ (Python), which the agent can run.
Shell commands in SKILL.md call:
python3gituvFrom the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md. Its commands use git and uv, which can reach the network depending on how they are called.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Dual Agent Collaboration loads about 2.2k tokens when it runs, and up to ~3.3k if it reads all its reference files. Until then it costs about 55 tokens; SKILL.md has 569 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.
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); the scripts in this folder are not scanned.
The full file from yunshu0909/yunshu_skillshub at commit da0d31f, republished under its MIT licence (© yunshu0909). 569 words, ~2,225 tokens.
.claude/skills/dual-agent-collaboration/SKILL.md (or your agent's skills folder). This skill also uses 7 other files; get the full folder from GitHub.把当前模型设为 primary,把另一个 CLI 模型设为 peer reviewer。Primary 对结果负责;peer 提供独立判断,不共同编辑同一工作树。
核心不变量:
先验证“做的是不是用户真正要的”,再验证“方案和代码是否正确”。两个模型对同一份错误合同达成一致,不算成功。
这里的“协作”默认指:primary 完成研究、实现与自测,peer 读取同一批原始证据做独立挑战和校核,primary 再按 finding 修正。Peer 不在共享工作树直接编码;这正是本 Skill 对既有“Codex 主做、Claude Code 独立校核”流程的通用化。若用户明确要双模型分工编码,必须使用独立 worktree,并把合并审查另列为任务,不能复用默认模式偷偷并行写。
“所有产品都能使用”指 Skill 可跨项目发现和运行,不表示每个琐碎任务都强制双模型。按比例触发:
一旦触发本 Skill,就完整走 Intent、Plan、Implementation、Final 四个 gate,不以“提速”为由省略。高效率来自:
默认 max reasoning 是完整协作的质量基线;只有用户明确优先成本/时延时才用环境变量下调,并在最终结果中披露。30 分钟是失败上限,不是预期耗时。
| 当前启动者 | Peer | 调用参数 |
|---|---|---|
| Codex | Claude Code 最新 Opus | --peer claude |
| Claude Code | Codex 当前高质量配置 | --peer codex |
| 无法确定 | 让脚本检测;检测冲突则显式指定 | --peer auto |
使用本 Skill 的 scripts/invoke_peer.py。脚本使用 argv + stdin 调 CLI,不使用 shell 拼接;peer 运行在只读模式,并设置递归保护。
先做命令构造预检:
python3 <skill-dir>/scripts/invoke_peer.py \
--peer auto \
--phase intent \
--cwd <project-root> \
--task-id <stable-project-task-id> \
--add-dir <external-evidence-dir> \
--prompt-file <handoff.md> \
--dry-run若 CLI 不存在、未认证、超时或未返回结构化结论,视为 BLOCKED,禁止假装已经双模型审查。
--dry-run 只证明参数已组装,不能证明 CLI 能启动、权限正确或输出可解析。任何 gate 的完成证据必须来自真实调用。
推荐用 --task-id,由 wrapper 按“项目路径 hash + task id”把 state 放入持久的用户状态目录:macOS 使用 ~/Library/Application Support/CodePal/dual-agent-workflows/,Linux 使用 $XDG_STATE_HOME/codepal/dual-agent-workflows/ 或 ~/.local/state/...。同一任务所有 gate 复用稳定 task id;不同任务不得复用。这样不污染仓库,也不会因系统清理 /tmp 或重启而丢掉 gate。只有集成方已有自己的状态存储时才显式用 --state-file。State 只存 verdict、session ID 和报告 hash,不存 handoff 正文。
读取 references/handoff-template.md,为每个门禁提供一份自包含 handoff。必须包含:
<raw_user_request> 标记,必须逐字引用,不能只给 primary 的总结。把绝对路径或仓库内路径交给 peer,让它直接读原始材料。Artifact manifest 至少包含一个真实存在的绝对路径。不要只贴结论,也不要在 handoff 中暗示“预期 ACK”或泄露你怀疑的 bug。
发送证据前先检查项目根的 AGENTS.md、CLAUDE.md、安全/隐私文档和用户约束:
Claude 的读取边界默认是 --cwd。任何证据在 cwd 外时,必须显式传入其父目录:
--add-dir /absolute/read-only/evidence-directoryClaude 仍只开放 Read/Grep/Glob,所以 add-dir 不授予写能力。Peer 如果读不到 manifest 中的材料,必须 BLOCKED;不得忽略后继续 ACK。
Codex 的 read-only sandbox 限制写入但允许读取 manifest 中的绝对路径,因此没有 --add-dir 参数。首次安装或 Codex CLI 版本变化时,必须用 cwd 外临时证据做一次真实读取探针;读不到就 BLOCKED,不得假设行为不变。
Wrapper 会强制检查四段机器标记:
<raw_user_request>...</raw_user_request>
<artifact_manifest>...</artifact_manifest>
<validation_evidence>...</validation_evidence>
<known_gaps>...</known_gaps>标记只能防遗漏,不能判断内容是否诚实;primary 仍必须保证其中是原始事实。
以下任一缺失时不得获得最终 ACK:
在设计或编码前,让 peer 独立回答:
python3 <skill-dir>/scripts/invoke_peer.py \
--peer auto --phase intent --cwd <project-root> \
--task-id <stable-project-task-id> \
--prompt-file <intent-handoff.md>NOT_ACK 时先修正需求合同。不要开始实现。若 primary 不同意 finding,用原始证据反驳并续审,不要因 peer 权威而盲从。
Intent ACK 后,把已接受的需求合同、技术调研和完整计划交给同一 peer session 续审:
python3 <skill-dir>/scripts/invoke_peer.py \
--peer <peer> --phase plan --cwd <project-root> \
--task-id <stable-project-task-id> \
--session-id <intent-session-id> \
--prompt-file <plan-handoff.md>Plan 必须覆盖正常路径、失败路径、迁移/回滚、隐私、安全、性能、真实数据验证和测试矩阵。Peer 必须重新对照用户原话,不能只检查计划内部自洽。
Primary 独立编码、测试和记录证据。Peer 默认不修改文件,避免共享工作树冲突和责任模糊。Claude peer 只开放 Read/Grep/Glob;Codex peer 使用 sandbox_mode=read-only + approval_policy=never。需要 git diff 时由 primary 预先生成只读文件并加入 artifact manifest。
实现完成后开启一个新的冷审 session,不要续接 Plan session,降低锚定效应。Handoff 必须包含:
git diff 可读路径;python3 <skill-dir>/scripts/invoke_peer.py \
--peer auto --phase implementation --cwd <project-root> \
--task-id <stable-project-task-id> \
--add-dir <diff-or-evidence-dir> \
--prompt-file <implementation-handoff.md>Peer 按 P0/P1/P2/note 报告 finding。Primary 修复后,用该 implementation session ID 续审,逐项关闭旧 finding,同时要求 fresh scan,不能只确认补丁表面存在。
只有以下条件全部满足才进入 final:
高风险任务、需求中途变化或真实验收推翻过方案时,Final 使用新的冷审 session;否则可续接 implementation session。
Wrapper 用 --task-id 自动定位 state(或由集成方显式传 --state-file),机器执行 gate 顺序:
intent 创建 workflow。plan 只接受已 ACK 的 intent,并必须续接 intent session。implementation 必须是新冷审 session;修复轮必须续接该 implementation session。final 只接受 intent / plan / implementation 全部 ACK。最终只接受三个 verdict:
ACK:合同和当前阶段证据充分,无阻断 finding。NOT_ACK:存在需要修复或重新定需求的问题。BLOCKED:缺少证据、权限、CLI 或用户决策,无法诚实判断。每轮遵循:
peer finding
→ primary 用证据复核
→ 接受并修复,或有证据地拒绝
→ primary 自测
→ resume 同一审查 session 逐项关闭
→ 要求 fresh scan
→ ACK / NOT_ACK / BLOCKED规则:
脚本参数:
--peer auto|claude|codex
--phase intent|plan|implementation|final
--cwd PROJECT_ROOT
--add-dir DIR # 可重复;Claude 读取 cwd 外证据时必传
--task-id ID # 推荐;按项目与任务自动隔离持久 state
--state-file PATH # 高级用法;与 --task-id 二选一
--prompt-file PATH # 与 --prompt / stdin 三选一
--prompt TEXT
--session-id ID # 续审
--timeout-seconds 1800
--save-raw DIR # 仅在确需保存完整 CLI 事件时使用
--dry-run输出是统一 JSON,至少包含:
{
"peer": "claude",
"phase": "implementation",
"sessionId": "...",
"verdict": "NOT_ACK",
"report": {
"summary": "...",
"findings": []
}
}默认不持久化原始 CLI 事件,因为其中可能包含项目内容。只有用户明确需要完整审查日志时才使用 --save-raw,并把目录放在安全位置。
模型选择:
opus alias + max effort,并启用 safe mode、禁用自定义 MCP、限制为 Read/Grep/Glob;macOS 上再用 sandbox-exec 禁止写 cwd/证据目录。用 DUAL_AGENT_CLAUDE_MODEL、DUAL_AGENT_CLAUDE_EFFORT 覆盖。max reasoning;用 DUAL_AGENT_CODEX_MODEL、DUAL_AGENT_CODEX_REASONING 覆盖。向用户汇报时说清:
若 primary 按比例原则判定某个低风险任务不触发本 Skill,必须向用户明确说明“本任务未启用双模型审查”及理由,不能让用户误以为已经发生协作。
首次安装或修改本 Skill 时,Final 还必须有这些可观察证据:
任何一项只有 dry-run、mock 或“理论上支持”,都不能写成已经完成。
Final handoff 必须把中央源目录和两个实际安装目录的绝对路径全部放进 <artifact_manifest>。Claude peer 审查时,还要为 cwd 外的每个安装目录传 --add-dir,让 peer 自己逐文件比较;任何目录不可读就 BLOCKED。
“三份一致”的唯一口径是:排除 __pycache__/、*.pyc、*.pyo、.DS_Store 后,比较全部相对文件路径集合和每个文件的 SHA-256;不得用目录大小、mtime 或抽样代替。
安装或更新采用可回滚顺序:
不要直接覆盖一个正在使用的目标目录,也不要在半成功状态下交付。
本 Skill 自带原子分发脚本,先 dry-run 再执行:
python3 <skill-dir>/scripts/install_skill.py \
--source <central-skill-dir> \
--target <codex-skill-dir> \
--target <claude-skill-dir> \
--dry-run移除 --dry-run 后才会 staging、比对、替换与回滚。脚本排除上述生成物,最终输出统一内容 manifest 的 SHA-256。
本 Skill 的官方结构检查命令是:
uv run --with pyyaml \
<skill-creator-dir>/scripts/quick_validate.py \
<dual-agent-collaboration-dir>预期输出包含 Skill is valid!;它只验证 Skill 结构,不替代真实 CLI 与新会话发现测试。
不得说“两个模型都同意,所以一定正确”。只能说:两套独立推理在完整证据上完成了交叉审查,当前没有已知阻断项。
© yunshu0909, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 7 other files (scripts, references) in coding/dual-agent-collaboration of yunshu0909/yunshu_skillshub.
Open the folder on GitHubat commit da0d31f
Dual Agent Collaboration 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 |
|---|---|---|---|---|---|---|
| Dual Agent Collaboration this skillyunshu0909/yunshu_skillshub | 768 | — | ~2.2k | Automated safety check: Pass | MIT | |
| User Story Writerdeanpeters/Product-Manager-Skills | 7.2k | 2 repos | ~2.9k | Automated safety check: Pass | Custom licence | |
| Game Changing FeaturesopenstatusHQ/data-table-filters | 2.3k | 3 repos | ~2.1k | Automated safety check: Pass | MIT | |
| CCPM Project Managementautomazeio/ccpm | 8.4k | — | ~1.1k | Automated safety check: Pass | MIT | |
| Convex Create Componentspokvulcan/poker-planning | 115 | 8 repos | ~2.6k | Automated safety check: Pass | MIT | |
| Self Improving Agentfarm-fe/farm | 5.6k | 2 repos | ~3.3k | Automated safety check: Notes | MIT |
deanpeters/Product-Manager-Skills
Writes user stories in Mike Cohn's format with Gherkin acceptance criteria, turning user needs into development-ready work with testable conditions.
openstatusHQ/data-table-filters
Find 10x product opportunities and high-leverage improvements.
automazeio/ccpm
Runs a spec-driven workflow from PRD to epic to GitHub issues to parallel agents, with status, standup and blocked-work reports from bundled scripts.
spokvulcan/poker-planning
Builds reusable Convex components with isolated tables and app-facing APIs.
farm-fe/farm
A universal self-improving agent that learns from ALL skill experiences.
garrytan/gstack
Builds a weekly engineering retrospective from git history: commit counts, per-person contributions, work patterns and code quality numbers over a chosen window.
yunshu0909/yunshu_skillshub
配图助手 - 把文章/模块内容转成统一风格、少字高可读的 16:9 信息图提示词;先定“需要几张图+每张讲什么”,再压缩文案与隐喻,最后输出可直接复制的生图提示词并迭代。
yunshu0909/yunshu_skillshub
和用户一起敲定一个前端页面的整页方案(三层法:先定这页给用户什么,再定含哪几样怎么排,最后才是长什么样),最后打成一个定稿包(含高保真全状态与交互流程图)交给开发。触发:用户对一个页面说「想重做 / 重新设计 / 整页不对 / 太满 / 说不清具体改哪」,或「我们讨论一下这页怎么设计」。覆盖三种情形:① 已有页面整页重做;②「替换」——新方案占用同一个入口(同一路由 /…
yunshu0909/yunshu_skillshub
和用户一起把一个功能「看不见的部分」敲定:数据存哪、谁写谁读、怎么加载、怎么保存、默认值和升级、核心执行规则、出问题时怎么办、旧东西怎么迁。每条规则带编号、案例和状态(已定 / 方案里有但你没拍板 / 待你定 / AI 定),用户只回答带案例的新选择题、过目没拍过板的规则;给用户看的是按「你会问的问题」分组的问答卡片(最绕的逻辑配可点的模拟器);最后请另一家模型只读查边界,问题闭环后封成后端逻辑…
yunshu0909/yunshu_skillshub
带用户精读一篇文章/文档并真正学透(不是出摘要)。五步:抽干货 → 切讲次 → 每讲跑「学-考-讲」循环 → 对号入座 → 实操+测验+讲错题+蒸馏。每讲产出 HTML 课件 + 笔记落盘;讲完必考一次,用户复述后必须挑不精确处拧紧;抽象概念上可交互演示(能点能跑);全程用用户自己的业务场景当案例;学完出多题型自动判分测验卷,最后把收获蒸馏回用户的工具。核心触发条件是用户要的是"学会"而不是"要…
yunshu0909/yunshu_skillshub
Issue 池全生命周期管理(开发范式 v1 规划段)。核心是一条 issue 驱动的流程:用户随手丢想法,你把糊的 issue 变成能开工的 task——产出的是“问题定义”,不是“解决方案实现”;载体就是仓库根的 ISSUES.md。五个动作:记、并、拆、转、pending。规划规模只区分 single-task(一个版本能交付)与 roadmap(需要多批滚动),不得使用…
yunshu0909/yunshu_skillshub
把项目推送到 GitHub,三种模式自动判断:首次推送(大文件扫描 → 生成 .gitignore → git init → gh 建仓 → 推送)、日常更新(commit + push)、版本发布(打 tag + 建 Release,可附下载文件)。核心原则是安全第一:推之前必扫大文件和敏感文件,宁可多问一句也不把不该推的东西推上去。当用户说"推到GitHub""推送到GitHub""git…
Categories
让 Codex 与 Claude Code 通过本机 CLI 组成“主执行者 + 独立审查者”的对称协作闭环。用于用户要求两个模型一起完成、交叉校核、独立审查、修到 ACK,或任务涉及产品需求收敛、复杂方案、跨模块开发、迁移、安全、重要重构和高质量交付时;无论从 Codex 还是 Claude Code 启动,都由当前模型主持,并完整调用另一方完成需求挑战、方案门禁、实现冷审和最终验收。. Dual Agent Collaboration is an agent skill from yunshu0909/yunshu_skillshub.
Dual Agent Collaboration fits situations like: product & Project Management work in your project.
Run `npx skills add yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a claude-code`. Or copy the skill folder (coding/dual-agent-collaboration in yunshu0909/yunshu_skillshub) into .claude/skills/dual-agent-collaboration in your project. Claude Code loads it when a task matches its description.
Run `npx skills add yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a codex`. Or copy the skill folder (coding/dual-agent-collaboration in yunshu0909/yunshu_skillshub) into .agents/skills/dual-agent-collaboration in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add yunshu0909/yunshu_skillshub --skill dual-agent-collaboration -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/dual-agent-collaboration, .gemini/skills/dual-agent-collaboration, .github/skills/dual-agent-collaboration and .opencode/skills/dual-agent-collaboration in your project.
Going by SKILL.md and its folder, Dual Agent Collaboration needs Python for the scripts in its folder and the command-line tools its instructions call (python3, git and uv). Our summary lists: Python 3.
SKILL.md contains no URLs. Its commands use git and uv, which can reach the network depending on how they are called. This is read from the text; nothing was executed.
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. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
Dual Agent Collaboration is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 2.2k tokens (SKILL.md is roughly 8.9k 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 1k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with Dual Agent Collaboration: User Story Writer (deanpeters/Product-Manager-Skills, 7.2k stars), Game Changing Features (openstatusHQ/data-table-filters, 2.3k stars), CCPM Project Management (automazeio/ccpm, 8.4k stars) and Convex Create Component (spokvulcan/poker-planning, 115 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
yunshu0909 (a GitHub user) maintains it in yunshu0909/yunshu_skillshub, which has 768 GitHub stars. The repository holds 25 skills in this directory. The repository was last updated on October 6, 2026.
Source: yunshu0909/yunshu_skillshub on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.