# Send 任务收尾器 to your agent Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually. ## Fast path - Download the package from Yavira. - Extract it into a folder your agent can access. - Paste one of the prompts below and point your agent at the extracted folder. ## Suggested prompts ### New install ```text I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Tell me what you changed and call out any manual steps you could not complete. ``` ### Upgrade existing ```text I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Summarize what changed and any follow-up checks I should run. ``` ## Machine-readable fields ```json { "schemaVersion": "1.0", "item": { "slug": "ren-wu-shou-wei-qi", "name": "任务收尾器", "source": "tencent", "type": "skill", "category": "效率提升", "sourceUrl": "https://clawhub.ai/Dalomeve/ren-wu-shou-wei-qi", "canonicalUrl": "https://clawhub.ai/Dalomeve/ren-wu-shou-wei-qi", "targetPlatform": "OpenClaw" }, "install": { "downloadUrl": "/downloads/ren-wu-shou-wei-qi", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=ren-wu-shou-wei-qi", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "packageFormat": "ZIP package", "primaryDoc": "SKILL.md", "includedAssets": [ "SKILL.md" ], "downloadMode": "redirect", "sourceHealth": { "source": "tencent", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-04-30T16:55:25.780Z", "expiresAt": "2026-05-07T16:55:25.780Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=network", "contentDisposition": "attachment; filename=\"network-1.0.0.zip\"", "redirectLocation": null, "bodySnippet": null }, "scope": "source", "summary": "Source download looks usable.", "detail": "Yavira can redirect you to the upstream package for this source.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/ren-wu-shou-wei-qi" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] } }, "links": { "detailUrl": "https://openagent3.xyz/skills/ren-wu-shou-wei-qi", "downloadUrl": "https://openagent3.xyz/downloads/ren-wu-shou-wei-qi", "agentUrl": "https://openagent3.xyz/skills/ren-wu-shou-wei-qi/agent", "manifestUrl": "https://openagent3.xyz/skills/ren-wu-shou-wei-qi/agent.json", "briefUrl": "https://openagent3.xyz/skills/ren-wu-shou-wei-qi/agent.md" } } ``` ## Documentation ### 任务收尾器 强制任务收尾，确保每个任务都有明确的完成证据。防止任务中途停滞、无证据完成。 ### 为什么需要这个技能 问题: AI 代理经常： 任务中途停止，无解释 只输出计划，不执行 缺少明确的完成标准 没有证据工件 解决方案: 此技能强制执行： 每个实质性步骤输出目标/进度/下一步 完成时必须提供证据 多步骤任务必须有完成证明格式 ### 1. 任务启动 每个任务开始时输出： **目标**: 完成时是什么样子 **进度**: 已完成什么 **下一步**: 现在执行的一个具体行动 ### 2. 执行中更新 每个实质性步骤后更新进度： **进度更新**: - 已完成：步骤 1, 2 - 当前：执行步骤 3 - 阻塞：无（或具体阻塞 + 已尝试 + 最小解锁输入） ### 3. 完成证明格式 2 步以上的任务必须包含： **DONE_CHECKLIST**: - [ ] 项目 1 已完成 - [ ] 项目 2 已完成 **EVIDENCE**: - 已执行：命令/操作摘要 - 工件：路径/URL/ID - 已验证：检查命令结果 **NEXT_AUTONOMOUS_STEP**: - 一个无需用户输入即可执行的后续步骤 ### 4. 反停滞规则 仅计划的回复：最多 1 次 下一次回复必须包含执行证据 永远不要以"我现在将..."结束而不显示工具结果 ### 可执行完成标准 任务完成当且仅当： 标准验证目标已陈述Select-String "目标" memory/{date}.md 匹配进度已追踪Select-String "进度" memory/{date}.md 匹配下一步已定义Select-String "下一步" memory/{date}.md 匹配证据存在工件路径/URL 存在无未解决标记Select-String "TODO ### 示例 1：发布技能到 ClawHub **目标**: 发布 skill 到 ClawHub，返回 URL 和 skill_id **进度**: - 已完成：读取 SKILL.md，打开浏览器 - 当前：填写发布表单 **下一步**: 填写 Slug/Name/Version 字段 --- **进度更新**: - 已完成：表单填写，文件上传 - 当前：等待发布确认 - 阻塞：无 --- **DONE_CHECKLIST**: - [x] 技能已发布 - [x] URL 已验证可访问 - [x] skill_id 已记录 **EVIDENCE**: - 已执行：clawhub publish 命令 - 工件：https://clawhub.ai/Dalomeve/my-skill - skill_id: k97xxxxx - 已验证：浏览器导航到 URL，200 OK **NEXT_AUTONOMOUS_STEP**: - 更新 INDEX.md 添加新技能条目 ### 示例 2：研究 GitHub 热门项目 **目标**: 研究 12+ GitHub 热门项目，提炼 3-5 个痛点，选择一个方向实现 **进度**: - 已完成：浏览 GitHub Trending，记录 14 个项目 - 当前：分析痛点，选择方向 **下一步**: 编写研究报告到 outputs/ --- **DONE_CHECKLIST**: - [x] 14 个项目已分析 - [x] 5 个痛点已识别 - [x] 研究方向已选择（验证空白） - [x] 新项目已创建并发布 **EVIDENCE**: - 研究报告：outputs/2026-03-01-github-research.md - 新项目：https://github.com/Dalomeve/agent-audit-trail - 测试通过：python test_audit_trail.py (4/4 tests passed) - Commit: 03307bdc84abab4bd5e78ab51abcd271ca85a4ab **NEXT_AUTONOMOUS_STEP**: - 将新项目添加到长期任务追踪 ### 隐私/安全 证据中不包含敏感数据 工件路径使用相对路径或工作空间路径 任务日志中不包含凭据 ### 自触发 当以下情况时使用： 启动任何多步骤任务 中断后恢复 移交给另一个代理 用户要求"不要只给计划，直接执行" ### 与其他技能集成 task-finish-contract: 英文版本，概念相同 agent-audit-trail: 使用审计追踪记录证据 phoenix-loop: 从失败中学习，更新收尾模式 HEARTBEAT.md: 心跳检查任务收尾状态 ### 限制 需要代理自觉遵守 证据验证是尽力的（URL 可能过期） 不适用于单步骤简单任务 ### 参考 tasks/QUEUE.md - 任务队列格式 memory/tasks.md - 任务历史 skills/task-finish-contract/ - 英文版本 有始有终。用证据证明。 ## Trust - Source: tencent - Verification: Indexed source record - Publisher: Dalomeve - Version: 1.0.2 ## Source health - Status: healthy - Source download looks usable. - Yavira can redirect you to the upstream package for this source. - Health scope: source - Reason: direct_download_ok - Checked at: 2026-04-30T16:55:25.780Z - Expires at: 2026-05-07T16:55:25.780Z - Recommended action: Download for OpenClaw ## Links - [Detail page](https://openagent3.xyz/skills/ren-wu-shou-wei-qi) - [Send to Agent page](https://openagent3.xyz/skills/ren-wu-shou-wei-qi/agent) - [JSON manifest](https://openagent3.xyz/skills/ren-wu-shou-wei-qi/agent.json) - [Markdown brief](https://openagent3.xyz/skills/ren-wu-shou-wei-qi/agent.md) - [Download page](https://openagent3.xyz/downloads/ren-wu-shou-wei-qi)