@givolt/i3-composer-agent 2.1.1 → 2.2.1

package/evals/zh-CN/cases/i3-code-simplification.json

@@ -0,0 +1,46 @@
+{
+  "skill_name": "i3-code-simplification",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "这个函数行为正确但太复杂，请在不改变行为的前提下简化。",
+        "top_k": 1
+      },
+      {
+        "prompt": "重构这段逻辑，提高可读性和维护性，减少不必要复杂度。",
+        "top_k": 1
+      },
+      {
+        "prompt": "请简化现有代码结构，保留全部现有行为。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "为项目制定覆盖率、性能和无障碍质量门槛，写入 CONSTRAINTS.md。",
+        "owner": "i3-constraint-driven-development"
+      },
+      {
+        "prompt": "昨天通过的测试现在失败，请复现并定位根因。",
+        "owner": "i3-debugging-and-error-recovery"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "简化提供的 80 行配置解析函数，精确保留行为。",
+      "expected_output": "A simpler implementation with identical behavior and a summary of what was removed and why",
+      "files": [
+        "i3-code-simplification"
+      ],
+      "expectations": [
+        "Behavior is preserved (tests unchanged and passing, or equivalence argued concretely)",
+        "Complexity is reduced rather than relocated",
+        "The response explains what was removed and why it was safe",
+        "No new features are added during the simplification",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-constraint-driven-development.json

@@ -0,0 +1,70 @@
+{
+  "skill_name": "i3-constraint-driven-development",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "为项目制定覆盖率、性能和无障碍质量门槛，写入 CONSTRAINTS.md。",
+        "top_k": 1
+      },
+      {
+        "prompt": "代理总是跳过测试和削弱断言，请建立不能随意降低的质量约束。",
+        "top_k": 1
+      },
+      {
+        "prompt": "我们没有具体数字，请访谈并推荐合理的项目质量标准。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "代理新会话忘记项目约定，请优化上下文和规则文件。",
+        "owner": "i3-context-engineering"
+      },
+      {
+        "prompt": "计划废弃旧版 API，让外部用户无中断迁移到新版。",
+        "owner": "i3-deprecation-and-migration"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "kind": "dialogue",
+      "prompt": "为这个仓库设置质量约束，我不知道应选择哪些数值。",
+      "expected_output": "A short interview with defaults offered, followed by a CONSTRAINTS.md the user can explain",
+      "expectations": [
+        "The agent inspects the repo for stack, test runner, and existing linters before asking anything",
+        "No more than four questions are asked, and each one carries a usable default",
+        "When the user has no target number, the agent proposes recording today's value and holding it rather than inventing a threshold",
+        "The resulting constraints say where each check runs, keeping fast checks in the edit loop and slow checks later",
+        "Every number written down is accompanied by a reason",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    },
+    {
+      "id": 1002,
+      "kind": "dialogue",
+      "prompt": "现有覆盖率为 62%，我希望从今天开始强制执行 85%。",
+      "expected_output": "Pushback on setting an unmeetable budget, with the ratchet alternative offered",
+      "expectations": [
+        "The agent points out that a threshold the codebase fails today produces a permanently red build",
+        "It offers to record 62% and require that it not fall, as an adoptable alternative",
+        "It does not silently accept 85% without noting the consequence",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    },
+    {
+      "id": 1003,
+      "kind": "dialogue",
+      "prompt": "无障碍对我们很重要，请把它加为约束。",
+      "expected_output": "A working accessibility check, not just a threshold written down",
+      "expectations": [
+        "The agent names axe-core and gives the install and run commands rather than only stating a violation threshold",
+        "It notes that axe needs a running URL and asks for one, or drops the dimension if the project has no URL to hit",
+        "The command that produces the verdict is recorded alongside the rule in CONSTRAINTS.md",
+        "The check is placed at the preview-deploy stage rather than in the per-edit loop",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-context-engineering.json

@@ -0,0 +1,45 @@
+{
+  "skill_name": "i3-context-engineering",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "代理新会话忘记项目约定，请优化上下文和规则文件。",
+        "top_k": 1
+      },
+      {
+        "prompt": "回复越来越泛化，帮我修复上下文配置和任务切换方式。",
+        "top_k": 1
+      },
+      {
+        "prompt": "如何组织项目规则和代理上下文以避免会话质量下降？",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "昨天通过的测试现在失败，请复现并定位根因。",
+        "owner": "i3-debugging-and-error-recovery"
+      },
+      {
+        "prompt": "把订单服务采用事件溯源的架构决策写成 ADR。",
+        "owner": "i3-documentation-and-adrs"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "代理会话质量下降：回复泛化，反复忘记项目约定。请修复上下文配置。",
+      "expected_output": "A diagnosis of the context problem and a concrete loading strategy (rules files, packing, task-scoped context)",
+      "files": [
+        "i3-context-engineering"
+      ],
+      "expectations": [
+        "The response diagnoses probable context causes rather than blaming the model generally",
+        "A specific context-loading plan is produced (what to load, when, and why)",
+        "Context is scoped to the task instead of loading everything",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-debugging-and-error-recovery.json

@@ -0,0 +1,60 @@
+{
+  "skill_name": "i3-debugging-and-error-recovery",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "昨天通过的测试现在失败，请复现并定位根因。",
+        "top_k": 1
+      },
+      {
+        "prompt": "应用突然崩溃，出现运行异常，帮我排查并验证修复。",
+        "top_k": 1
+      },
+      {
+        "prompt": "构建失败和报错需要系统诊断，请查清哪里坏了。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "计划废弃旧版 API，让外部用户无中断迁移到新版。",
+        "owner": "i3-deprecation-and-migration"
+      },
+      {
+        "prompt": "在不可逆数据迁移前，对计划进行对抗式审查和假设核查。",
+        "owner": "i3-doubt-driven-development"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "昨天通过的分页测试今天出现差一错误，请查明并修复根因。",
+      "expected_output": "A reproduced failure, an identified root cause, a minimal fix, and a guard against regression",
+      "files": [
+        "i3-debugging-and-error-recovery"
+      ],
+      "expectations": [
+        "The failure is reproduced before any fix is attempted",
+        "The root cause is identified and stated, not just the symptom patched",
+        "A regression test exists after the fix",
+        "The fix is minimal and scoped to the cause",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    },
+    {
+      "id": 1002,
+      "prompt": "生产结账功能故障，利益相关者要求不复现就立即打分页补丁。请在时间压力下作出响应。",
+      "expected_output": "Evidence preserved, the failure reproduced, and a root-cause fix guarded by a regression test",
+      "files": [
+        "i3-debugging-and-error-recovery"
+      ],
+      "expectations": [
+        "The time pressure does not cause reproduction or evidence gathering to be skipped",
+        "The requested speculative patch is not applied before the cause is localized",
+        "The final fix is tied to a reproduced cause and a regression test",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-deprecation-and-migration.json

@@ -0,0 +1,45 @@
+{
+  "skill_name": "i3-deprecation-and-migration",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "计划废弃旧版 API，让外部用户无中断迁移到新版。",
+        "top_k": 1
+      },
+      {
+        "prompt": "生产数据库如何用扩展收缩方式无停机重命名列？",
+        "top_k": 1
+      },
+      {
+        "prompt": "迁移用户后如何退役旧系统并安排废弃周期？",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "把订单服务采用事件溯源的架构决策写成 ADR。",
+        "owner": "i3-documentation-and-adrs"
+      },
+      {
+        "prompt": "开发一个响应式前端页面，做好组件布局和无障碍交互。",
+        "owner": "i3-frontend-ui-engineering"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "规划公共 v1 REST API 的废弃，现有 200 个外部调用方，使用 v2 替代。",
+      "expected_output": "A staged deprecation plan: comms, timeline, compatibility window, monitoring, and removal criteria",
+      "files": [
+        "i3-deprecation-and-migration"
+      ],
+      "expectations": [
+        "Consumers are notified with a timeline before any breaking change",
+        "A compatibility or migration window exists with monitoring of remaining usage",
+        "Removal is gated on measured migration, not a calendar date alone",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-documentation-and-adrs.json

@@ -0,0 +1,45 @@
+{
+  "skill_name": "i3-documentation-and-adrs",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "把订单服务采用事件溯源的架构决策写成 ADR。",
+        "top_k": 1
+      },
+      {
+        "prompt": "公共 API 发生变化，请补充文档并记录设计理由。",
+        "top_k": 1
+      },
+      {
+        "prompt": "保存这次技术决策的背景与取舍，方便后续工程师理解。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "在不可逆数据迁移前，对计划进行对抗式审查和假设核查。",
+        "owner": "i3-doubt-driven-development"
+      },
+      {
+        "prompt": "工作区混着修复、重构和新功能，请拆成清晰的 Git 提交。",
+        "owner": "i3-git-workflow-and-versioning"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "把订单服务采用事件溯源的决策记录为 ADR。",
+      "expected_output": "An ADR capturing context, decision, alternatives considered, and consequences",
+      "files": [
+        "i3-documentation-and-adrs"
+      ],
+      "expectations": [
+        "The ADR states context, decision, alternatives, and consequences distinctly",
+        "Trade-offs and rejected options are recorded, not just the winning choice",
+        "The ADR records a status from the skill's lifecycle (Proposed, Accepted, Superseded or Deprecated) and a date",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-doubt-driven-development.json

@@ -0,0 +1,45 @@
+{
+  "skill_name": "i3-doubt-driven-development",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "在不可逆数据迁移前，对计划进行对抗式审查和假设核查。",
+        "top_k": 1
+      },
+      {
+        "prompt": "请独立质询每个非平凡决策，找出隐藏失败模式。",
+        "top_k": 1
+      },
+      {
+        "prompt": "这是安全敏感的生产认证改动，请对方案做压力测试和反证审查。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "开发一个响应式前端页面，做好组件布局和无障碍交互。",
+        "owner": "i3-frontend-ui-engineering"
+      },
+      {
+        "prompt": "这个团队知识共享工具的想法还很模糊，帮我发散并收敛方向。",
+        "owner": "i3-idea-refine"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "执行不可逆数据迁移前，对迁移计划进行对抗式审查。",
+      "expected_output": "Claims extracted, doubts raised against each, reconciliation, and a go or stop verdict",
+      "files": [
+        "i3-doubt-driven-development"
+      ],
+      "expectations": [
+        "Non-trivial claims in the plan are extracted and challenged individually",
+        "At least one assumption is tested rather than accepted",
+        "The verdict distinguishes verified claims from surviving doubts",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-frontend-ui-engineering.json

@@ -0,0 +1,45 @@
+{
+  "skill_name": "i3-frontend-ui-engineering",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "开发一个响应式前端页面，做好组件布局和无障碍交互。",
+        "top_k": 1
+      },
+      {
+        "prompt": "实现设计系统的下拉菜单，兼顾键盘操作和状态管理。",
+        "top_k": 1
+      },
+      {
+        "prompt": "请调整移动端界面布局和组件样式，达到生产视觉质量。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "工作区混着修复、重构和新功能，请拆成清晰的 Git 提交。",
+        "owner": "i3-git-workflow-and-versioning"
+      },
+      {
+        "prompt": "按照现有任务计划增量实现 CSV 导出，每个切片都验证。",
+        "owner": "i3-incremental-implementation"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "为设计系统实现一个下拉菜单组件。",
+      "expected_output": "An accessible, keyboard-navigable component following project conventions",
+      "files": [
+        "i3-frontend-ui-engineering"
+      ],
+      "expectations": [
+        "Keyboard interaction and focus management are implemented, not just mouse clicks",
+        "ARIA roles or semantic elements are used correctly",
+        "Component state is managed deliberately rather than ad hoc",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-git-workflow-and-versioning.json

@@ -0,0 +1,45 @@
+{
+  "skill_name": "i3-git-workflow-and-versioning",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "工作区混着修复、重构和新功能，请拆成清晰的 Git 提交。",
+        "top_k": 1
+      },
+      {
+        "prompt": "如何组织分支、解决合并冲突并创建 PR？",
+        "top_k": 1
+      },
+      {
+        "prompt": "准备发行版本，请安排语义化版本、标签和变更日志。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "这个团队知识共享工具的想法还很模糊，帮我发散并收敛方向。",
+        "owner": "i3-idea-refine"
+      },
+      {
+        "prompt": "我只知道想要更好的管理页面，请先一次一个问题访谈我。",
+        "owner": "i3-interview-me"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "工作区混着一次重构、一个缺陷修复和一个新功能，请整理成清晰的提交历史。",
+      "expected_output": "Separate atomic commits with clear messages, each independently green",
+      "files": [
+        "i3-git-workflow-and-versioning"
+      ],
+      "expectations": [
+        "Refactor, fix, and feature land as separate commits",
+        "Commit messages are imperative and standalone",
+        "Each commit leaves the tree in a working state",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-idea-refine.json

@@ -0,0 +1,44 @@
+{
+  "skill_name": "i3-idea-refine",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "这个团队知识共享工具的想法还很模糊，帮我发散并收敛方向。",
+        "top_k": 1
+      },
+      {
+        "prompt": "请细化这个创意，验证假设并比较更多选项。",
+        "top_k": 1
+      },
+      {
+        "prompt": "我想先完善想法，再确定实施方向。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "按照现有任务计划增量实现 CSV 导出，每个切片都验证。",
+        "owner": "i3-incremental-implementation"
+      },
+      {
+        "prompt": "给支付重试增加日志、指标、链路追踪和告警，便于值班。",
+        "owner": "i3-observability-and-instrumentation"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "kind": "dialogue",
+      "prompt": "完善这个模糊想法：某种帮助团队更好共享知识的工具。",
+      "expected_output": "Sharpening questions, distinct directions, surfaced assumptions, and a one-pager with MVP scope and a Not Doing list",
+      "expectations": [
+        "Sharpening questions are asked before converging",
+        "Hidden assumptions are surfaced explicitly",
+        "The output includes an explicit Not Doing list",
+        "The agent pushes back on weak aspects instead of only agreeing",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-incremental-implementation.json

@@ -0,0 +1,59 @@
+{
+  "skill_name": "i3-incremental-implementation",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "按照现有任务计划增量实现 CSV 导出，每个切片都验证。",
+        "top_k": 1
+      },
+      {
+        "prompt": "这个功能涉及多个文件，请拆成完整的垂直实现切片。",
+        "top_k": 1
+      },
+      {
+        "prompt": "领取计划的下一任务，小步实现、测试并提交。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "我只知道想要更好的管理页面，请先一次一个问题访谈我。",
+        "owner": "i3-interview-me"
+      },
+      {
+        "prompt": "产品页面有一千条数据渲染很慢，请定位性能瓶颈并优化。",
+        "owner": "i3-performance-optimization"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "按已有任务计划，为报表页面实现 CSV 导出。",
+      "expected_output": "The feature delivered in small verified increments with a commit per slice",
+      "files": [
+        "i3-incremental-implementation"
+      ],
+      "expectations": [
+        "Work proceeds in thin vertical slices rather than one large change",
+        "Each slice is verified (tests or build) before the next begins",
+        "Each slice is committed separately",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    },
+    {
+      "id": 1002,
+      "prompt": "两天的 CSV 导出草稿混合所有层且没有测试，管理层说拆分会浪费已投入工作。请决定如何继续。",
+      "expected_output": "The draft is converted into small independently verified slices without accepting sunk cost as a reason to batch",
+      "files": [
+        "incremental-implementation-pressure"
+      ],
+      "expectations": [
+        "Sunk cost is not accepted as a reason to commit the unverified batch",
+        "The work is decomposed into independently useful vertical slices",
+        "Verification is required before each slice is committed",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-interview-me.json

@@ -0,0 +1,43 @@
+{
+  "skill_name": "i3-interview-me",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "我只知道想要更好的管理页面，请先一次一个问题访谈我。",
+        "top_k": 1
+      },
+      {
+        "prompt": "需求还缺少目标用户和目的，先质询并明确真实意图。",
+        "top_k": 1
+      },
+      {
+        "prompt": "不要先给方案，请采访我，直到你确信理解我想要什么。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "给支付重试增加日志、指标、链路追踪和告警，便于值班。",
+        "owner": "i3-observability-and-instrumentation"
+      },
+      {
+        "prompt": "把通知功能规格拆成有依赖顺序和验收条件的任务计划。",
+        "owner": "i3-planning-and-task-breakdown"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "kind": "dialogue",
+      "prompt": "我想要一个更好的管理页面。先访谈我，再提出任何建议。",
+      "expected_output": "A one-question-at-a-time interview that converges on validated requirements",
+      "expectations": [
+        "Questions are asked one at a time, not in batches",
+        "The agent does not propose solutions before understanding the need",
+        "The interview surfaces the underlying goal behind the stated ask",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}

package/evals/zh-CN/cases/i3-observability-and-instrumentation.json

@@ -0,0 +1,46 @@
+{
+  "skill_name": "i3-observability-and-instrumentation",
+  "trigger": {
+    "positive": [
+      {
+        "prompt": "给支付重试增加日志、指标、链路追踪和告警，便于值班。",
+        "top_k": 1
+      },
+      {
+        "prompt": "生产功能需要可观测性，请提供运行证据和监控方案。",
+        "top_k": 1
+      },
+      {
+        "prompt": "线上故障无法从已有日志定位，请补齐观测和埋点。",
+        "top_k": 1
+      }
+    ],
+    "negative": [
+      {
+        "prompt": "产品页面有一千条数据渲染很慢，请定位性能瓶颈并优化。",
+        "owner": "i3-performance-optimization"
+      },
+      {
+        "prompt": "加固接收用户 webhook URL 的端点，检查 SSRF 漏洞。",
+        "owner": "i3-security-and-hardening"
+      }
+    ]
+  },
+  "evals": [
+    {
+      "id": 1001,
+      "prompt": "为新的支付重试功能增加观测，让值班人员可以运营它。",
+      "expected_output": "On-call questions defined first, then structured logs, RED metrics, and symptom-based alerts that answer them",
+      "files": [
+        "i3-observability-and-instrumentation"
+      ],
+      "expectations": [
+        "On-call questions are written before instrumentation is added",
+        "Logs are structured events with a correlation id, not prose strings",
+        "Metrics avoid unbounded label cardinality",
+        "Alerts are symptom-based and actionable",
+        "人类可读的追问、说明、进度、报告与生成文档使用简体中文；保留代码、路径、工具参数、JSON 协议及机器状态的原始格式。"
+      ]
+    }
+  ]
+}