iomgaa
0fe1c96393
Complete v3 planning: construction-first paradigm (frame-perception grounded fact extraction + 4-family independent judges + 6-layer verification + QuestionUnit contract), adversarial audit, paradigm-shift finding, real-data spike validation, logging schema, and phase1 contract implementation plan.
384 lines
12 KiB
JSON
384 lines
12 KiB
JSON
{
|
||
"directed": true,
|
||
"nodes": [
|
||
{
|
||
"id": "plan:infrastructure-setup",
|
||
"label": "项目基础设施初始化计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:core-agent-adapters-llm",
|
||
"label": "core/agent/ + adapters/llm 基础设施设计",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:core-agent-adapters-llm",
|
||
"label": "core/agent/ + adapters/llm 基础设施实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:tree-module-vertical-slice",
|
||
"label": "建树模块竖切设计：数据结构 + 建树 + 修复 + 迁移",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:tree-module-vertical-slice",
|
||
"label": "建树模块竖切实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "plan:question-gen",
|
||
"label": "question_gen 模块实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:2026-07-07-search-module-design",
|
||
"label": "搜索 Agent 装配层设计（app/search/）",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:2026-07-07-search-module",
|
||
"label": "app/search/ 搜索 Agent 装配层实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "plan:app-harness",
|
||
"label": "app/harness/ 训练循环编排层实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:tree-repair-resilience",
|
||
"label": "建树修复管线：熔断根因修复 + 断点续跑 + 并发改造",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:tree-repair-resilience",
|
||
"label": "建树修复管线三项改造实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:paper-main-figure",
|
||
"label": "论文主图：Self-Evolving Search Agent 推理训练闭环",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "design:question-gen-synth",
|
||
"label": "赛题生成工具设计",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:question-gen-synth",
|
||
"label": "赛题生成工具实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:main-inference-entry",
|
||
"label": "main.py 推理入口 + 初始 Prompt 集设计",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:main-inference-entry",
|
||
"label": "main.py 推理入口 + 初始 Prompt 集实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:agent-runtime-fixes",
|
||
"label": "Spec-1 Agent 执行环境修复（解析容错+步级重试+摘要附实体）",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "design:batch-tree-build",
|
||
"label": "Spec-2 建树批量并行入口",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "design:question-gen-v2",
|
||
"label": "Spec-3 出题管线 v2（失败机理靶向+逐题质量门）",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:agent-runtime-fixes-plan",
|
||
"label": "Spec-1 Agent 执行环境修复实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "finding:eval-spec1check",
|
||
"label": "Harness 评估: Spec-1 修复验证 (infer_spec1check)",
|
||
"type": "finding"
|
||
},
|
||
{
|
||
"id": "plan:batch-tree-build-plan",
|
||
"label": "Spec-2 建树批量并行实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "schema:question-gen-runs",
|
||
"label": "表结构: question_gen_runs（出题批次）",
|
||
"type": "schema"
|
||
},
|
||
{
|
||
"id": "schema:question-gen-items",
|
||
"label": "表结构: question_gen_items（逐题门判定）",
|
||
"type": "schema"
|
||
},
|
||
{
|
||
"id": "metric:qgen-gate-rejection-rate",
|
||
"label": "出题四门总拦截率",
|
||
"type": "metric"
|
||
},
|
||
{
|
||
"id": "metric:qgen-regen-convergence",
|
||
"label": "重出收敛率（3轮内过门）",
|
||
"type": "metric"
|
||
},
|
||
{
|
||
"id": "metric:qgen-acceptance-rate",
|
||
"label": "最终接受率",
|
||
"type": "metric"
|
||
},
|
||
{
|
||
"id": "metric:qgen-answer-uniformity",
|
||
"label": "答案位置分布均匀性",
|
||
"type": "metric"
|
||
},
|
||
{
|
||
"id": "plan:2026-07-11-question-gen-v2",
|
||
"label": "出题管线 v2 实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:per-category-pool-strategy",
|
||
"label": "Per-Category Pool Strategy 设计",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:per-category-pool-strategy",
|
||
"label": "Per-Category Pool Strategy 实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "plan:action-recognition-training",
|
||
"label": "Action Recognition 单题型首次训练实验计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "plan:task-type-strategy-framework",
|
||
"label": "TaskTypeStrategy 框架实现计划 (Plan A)",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "plan:action-recognition-strategy",
|
||
"label": "ActionRecognitionStrategy 特化实现计划 (Plan B)",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "design:maintenance-pool",
|
||
"label": "训练池 Maintenance 正确题自动补入机制",
|
||
"type": "design"
|
||
},
|
||
{
|
||
"id": "plan:maintenance-pool",
|
||
"label": "Maintenance Pool 自动补入实现计划",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "plan:grounded-question-gen-phaseA",
|
||
"label": "Grounded Question-Gen Phase A",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "plan:adversarial-question-gen-phaseB",
|
||
"label": "Adversarial Question-Gen Phase B",
|
||
"type": "plan"
|
||
},
|
||
{
|
||
"id": "plan:question-gen-v3-phase1-contract",
|
||
"label": "question-gen v3 Phase1 契约地基实现计划",
|
||
"type": "plan"
|
||
}
|
||
],
|
||
"links": [
|
||
{
|
||
"source": "plan:core-agent-adapters-llm",
|
||
"target": "design:core-agent-adapters-llm",
|
||
"relation": "implements",
|
||
"evidence": "计划实现设计文档中定义的全异步 AgentLoop + 四层治理栈",
|
||
"added": "2026-07-07T02:25:32.349931+00:00"
|
||
},
|
||
{
|
||
"source": "plan:tree-module-vertical-slice",
|
||
"target": "design:tree-module-vertical-slice",
|
||
"relation": "implements",
|
||
"evidence": "计划逐 Task 实现设计文档中的 11 个模块",
|
||
"added": "2026-07-07T05:27:02.953166+00:00"
|
||
},
|
||
{
|
||
"source": "plan:question-gen",
|
||
"target": "design:question-gen",
|
||
"relation": "implements",
|
||
"evidence": "实现计划覆盖设计文档的全部需求",
|
||
"added": "2026-07-07T08:32:53.887071+00:00"
|
||
},
|
||
{
|
||
"source": "plan:2026-07-07-search-module",
|
||
"target": "design:2026-07-07-search-module-design",
|
||
"relation": "implements",
|
||
"evidence": "实现搜索 Agent 装配层设计",
|
||
"added": "2026-07-07T09:36:21.467921+00:00"
|
||
},
|
||
{
|
||
"source": "plan:app-harness",
|
||
"target": "design:app-harness-design",
|
||
"relation": "implements",
|
||
"evidence": "实现 2026-07-07-app-harness-design.md 的 14 文件训练循环编排层",
|
||
"added": "2026-07-07T15:45:08.729979+00:00"
|
||
},
|
||
{
|
||
"source": "plan:tree-repair-resilience",
|
||
"target": "design:tree-repair-resilience",
|
||
"relation": "implements",
|
||
"evidence": "实现设计文档的三项改造：遥测加固+断点续跑+并发",
|
||
"added": "2026-07-09T04:08:15.312470+00:00"
|
||
},
|
||
{
|
||
"source": "plan:question-gen-synth",
|
||
"target": "design:question-gen-synth",
|
||
"relation": "implements",
|
||
"evidence": "计划实现设计文档中定义的 synthesizer + factory + CLI 三模块",
|
||
"added": "2026-07-09T09:05:43.697644+00:00"
|
||
},
|
||
{
|
||
"source": "plan:main-inference-entry",
|
||
"target": "design:main-inference-entry",
|
||
"relation": "implements",
|
||
"evidence": "实现设计文档中定义的 main.py + deps_router + skills/v1",
|
||
"added": "2026-07-09T14:43:06.084216+00:00"
|
||
},
|
||
{
|
||
"source": "plan:agent-runtime-fixes-plan",
|
||
"target": "design:agent-runtime-fixes",
|
||
"relation": "implements",
|
||
"evidence": "实现 Spec-1 三处改动 A1/A2/B",
|
||
"added": "2026-07-11T12:06:25.879643+00:00"
|
||
},
|
||
{
|
||
"source": "plan:batch-tree-build-plan",
|
||
"target": "design:batch-tree-build",
|
||
"relation": "implements",
|
||
"evidence": "实现 Spec-2 四任务",
|
||
"added": "2026-07-11T15:07:06.178664+00:00"
|
||
},
|
||
{
|
||
"source": "metric:qgen-gate-rejection-rate",
|
||
"target": "schema:question-gen-items",
|
||
"relation": "measures",
|
||
"evidence": "从 question_gen_items 表 attempt=1 行聚合计算",
|
||
"added": "2026-07-12T02:34:28.728290+00:00"
|
||
},
|
||
{
|
||
"source": "metric:qgen-regen-convergence",
|
||
"target": "schema:question-gen-items",
|
||
"relation": "measures",
|
||
"evidence": "按 slot_id 聚合 final_status 计算收敛率",
|
||
"added": "2026-07-12T02:34:29.858363+00:00"
|
||
},
|
||
{
|
||
"source": "metric:qgen-acceptance-rate",
|
||
"target": "schema:question-gen-runs",
|
||
"relation": "measures",
|
||
"evidence": "直接读取 runs 表 accepted/total_slots",
|
||
"added": "2026-07-12T02:34:30.989357+00:00"
|
||
},
|
||
{
|
||
"source": "schema:question-gen-runs",
|
||
"target": "design:2026-07-11-question-gen-v2",
|
||
"relation": "implements",
|
||
"evidence": "设计 §5 运行时数据要求",
|
||
"added": "2026-07-12T02:34:32.121484+00:00"
|
||
},
|
||
{
|
||
"source": "schema:question-gen-items",
|
||
"target": "design:2026-07-11-question-gen-v2",
|
||
"relation": "implements",
|
||
"evidence": "设计 §5 逐题可追溯要求",
|
||
"added": "2026-07-12T02:34:33.267570+00:00"
|
||
},
|
||
{
|
||
"source": "plan:2026-07-11-question-gen-v2",
|
||
"target": "design:2026-07-11-question-gen-v2",
|
||
"relation": "implements",
|
||
"evidence": "Spec-3 设计的实现计划",
|
||
"added": "2026-07-12T02:57:39.569968+00:00"
|
||
},
|
||
{
|
||
"source": "plan:per-category-pool-strategy",
|
||
"target": "design:per-category-pool-strategy",
|
||
"relation": "implements",
|
||
"evidence": "实现 PoolStrategy Protocol + PerCategoryPoolStrategy 设计",
|
||
"added": "2026-07-13T02:26:58.551190+00:00"
|
||
},
|
||
{
|
||
"source": "plan:action-recognition-training",
|
||
"target": "design:action-recognition-training",
|
||
"relation": "implements",
|
||
"evidence": "计划实现设计文档中的 2 处代码修改 + 实验配置 + 训练脚本",
|
||
"added": "2026-07-14T04:50:15.986586+00:00"
|
||
},
|
||
{
|
||
"source": "plan:task-type-strategy-framework",
|
||
"target": "design:task-type-strategy",
|
||
"relation": "implements",
|
||
"evidence": "Plan A 实现设计中的框架层（Protocol + Base + pipeline 集成）",
|
||
"added": "2026-07-14T09:15:22.951263+00:00"
|
||
},
|
||
{
|
||
"source": "plan:action-recognition-strategy",
|
||
"target": "design:task-type-strategy",
|
||
"relation": "implements",
|
||
"evidence": "Plan B implements design §4: ActionRecognitionStrategy with 6 SubPatterns",
|
||
"added": "2026-07-14T10:34:27.079088+00:00"
|
||
},
|
||
{
|
||
"source": "design:maintenance-pool",
|
||
"target": "design:task-type-strategy",
|
||
"relation": "extends",
|
||
"evidence": "maintenance pool supplements the per-category pool strategy to support SubPattern-targeted training",
|
||
"added": "2026-07-14T14:11:13.527027+00:00"
|
||
},
|
||
{
|
||
"source": "plan:maintenance-pool",
|
||
"target": "design:maintenance-pool",
|
||
"relation": "implements",
|
||
"evidence": "Implements maintenance pool auto-supplement in PerCategoryPoolStrategy",
|
||
"added": "2026-07-14T14:22:44.443012+00:00"
|
||
},
|
||
{
|
||
"source": "plan:grounded-question-gen-phaseA",
|
||
"target": "design:grounded-question-gen-phaseA",
|
||
"relation": "implements",
|
||
"evidence": "Phase A 实现计划",
|
||
"added": "2026-07-14T18:41:35.899803+00:00"
|
||
},
|
||
{
|
||
"source": "plan:adversarial-question-gen-phaseB",
|
||
"target": "design:adversarial-question-gen-phaseB",
|
||
"relation": "implements",
|
||
"evidence": "Phase B 实现计划",
|
||
"added": "2026-07-14T21:09:48.115521+00:00"
|
||
},
|
||
{
|
||
"source": "schema:v3-question-gen-logging",
|
||
"target": "design:question-gen-v3-construction-paradigm",
|
||
"relation": "implements",
|
||
"evidence": "v3 出题运行时数据 schema 实现设计 §12.1",
|
||
"added": "2026-07-15T09:18:41.713855+00:00"
|
||
},
|
||
{
|
||
"source": "plan:question-gen-v3-phase1-contract",
|
||
"target": "design:question-gen-v3-construction-paradigm",
|
||
"relation": "implements",
|
||
"evidence": "Phase1 契约地基",
|
||
"added": "2026-07-15T09:22:08.800078+00:00"
|
||
}
|
||
]
|
||
} |