方法论库索引
Z2 条目↑ Z1 版块 ↓ Z3 全文
44 篇跨项目可复用方法论 · T1 三联表分级 · 按领域分组
方法论库 44 篇跨项目可复用方法论，按 INDEX.md 领域分组。T1 三联表分级：实跑=11 / 引用级=32 / orphan=1；surface 分布：none=33 / memory_index=11。
领域研究方法蒸馏 (1)
| slug | 一句话 | T1 分级 | 真源路径 |
|---|---|---|---|
| knowledge_transformation_theories_20260407_manual | 信息到知识的七个认知/组织理论 | 引用级 | contexts/methodology/knowledge_transformation_theories_20260407_manual.md |
跨域综合 (6)
| slug | 一句话 | T1 分级 | 真源路径 |
|---|---|---|---|
| chaos_to_order_knowledge_distillation_synthesis_20260407_manual | 六领域同构模式综合 | 引用级 | contexts/methodology/chaos_to_order_knowledge_distillation_synthesis_20260407_manual.md |
| chaos_to_order_round2_synthesis_20260407_manual | 第二轮综合+失败案例批判 | 实跑 | contexts/methodology/chaos_to_order_round2_synthesis_20260407_manual.md |
| essence_cross_domain_methodology_20260511_deep_research | 本质同构发现五步法 | orphan | contexts/methodology/essence_cross_domain_methodology_20260511_deep_research.md |
| se_module_split_merge_history_20260703_manual → workflow_idea_to_production | 50 年软工模块拆合理论横向综合 | 实跑 | contexts/methodology/se_module_split_merge_history_20260703_manual.md |
| task_decomposition_history_methodologies_survey_20260416 | 军事/软工/组织史任务拆解横向综合 | 引用级 | contexts/methodology/task_decomposition_history_methodologies_survey_20260416.md |
| diagram_for_understanding_methodologies_20260627_manual | 四大制图传统的结构揭示方法 | 实跑 | contexts/methodology/diagram_for_understanding_methodologies_20260627_manual.md |
Agent/LLM 理论框架 (10)
| slug | 一句话 | T1 分级 | 真源路径 |
|---|---|---|---|
| agent_open_loop_control_theory_20260531_manual | 开环控制统一解释 agent 失效 | 实跑 memory_index | contexts/methodology/agent_open_loop_control_theory_20260531_manual.md |
| anti_performative_diligence_methodology_20260530_manual | 反表演式勤奋三杠杆 | 实跑 memory_index | contexts/methodology/anti_performative_diligence_methodology_20260530_manual.md |
| agent_native_data_paradigm_20260531_deep_research | LLM 原生数据范式 | 实跑 memory_index | contexts/methodology/agent_native_data_paradigm_20260531_deep_research.md |
| agent_task_gap_closure_theorem_20260522_manual | Agent/Task/Gap 闭合定理 | 引用级 | contexts/methodology/agent_task_gap_closure_theorem_20260522_manual.md |
| agentic_task_ontology_20260521_manual | Agentic Task 本体 | 引用级 | contexts/methodology/agentic_task_ontology_20260521_manual.md |
| agentic_task_realization_20260521_manual | Agentic Task 实现闭环 | 引用级 | contexts/methodology/agentic_task_realization_20260521_manual.md |
| llm_as_probabilistic_system_20260520_manual | LLM 作为概率系统的任务设计 | 引用级 | contexts/methodology/llm_as_probabilistic_system_20260520_manual.md |
| context_inertia_task_assignment_20260429_deep_research | 上下文惯性与任务分配 | 引用级 | contexts/methodology/context_inertia_task_assignment_20260429_deep_research.md |
| assumed_defect_reflection_interrupt_20260607_deep_research | 假设缺陷探针中断机制 | 引用级 | contexts/methodology/assumed_defect_reflection_interrupt_20260607_deep_research.md |
| agent_supervision_context_drift_prevention_20260407_manual | 大上下文 agent 监督机制编目 | 引用级 | contexts/methodology/agent_supervision_context_drift_prevention_20260407_manual.md |
LLM 能力边界与任务拆解 (13)
| slug | 一句话 | T1 分级 | 真源路径 |
|---|---|---|---|
| llm_capability_boundary_20260417_deep_research | L0-L3×S0-S3 能力边界框架 | 引用级 | contexts/methodology/llm_capability_boundary_20260417_deep_research.md |
| human_llm_boundary_20260416_deep_research | 推理/指令遵循能力综述 | 引用级 | contexts/methodology/human_llm_boundary_20260416_deep_research.md |
| human_llm_capability_boundary_survey_20260416 | 同主题跨轮综合 | 引用级 | contexts/methodology/human_llm_capability_boundary_survey_20260416.md |
| llm_constraint_capacity_20260414_deep_research | 约束容量基准量化 | 引用级 memory_index | contexts/methodology/llm_constraint_capacity_20260414_deep_research.md |
| llm_constraint_capacity_and_task_decomposition_survey_20260414 | 3-6 约束上限+工作记忆同构 | 引用级 memory_index | contexts/methodology/llm_constraint_capacity_and_task_decomposition_survey_20260414.md |
| llm_context_multi_topic_20260416_deep_research | 多主题上下文容量 | 引用级 | contexts/methodology/llm_context_multi_topic_20260416_deep_research.md |
| llm_context_multi_topic_capacity_survey_20260416 | 主题/约束预算模型 | 引用级 | contexts/methodology/llm_context_multi_topic_capacity_survey_20260416.md |
| llm_task_decomposition_20260414_deep_research | 任务拆解 claim 验证 | 引用级 memory_index | contexts/methodology/llm_task_decomposition_20260414_deep_research.md |
| llm_task_decomposition_survey_20260301_deep_research | 拆解技术综述（CoT/ToT/ReAct） | 引用级 | contexts/methodology/llm_task_decomposition_survey_20260301_deep_research.md |
| task_decomposition_thinking_survey_20260303_deep_research | 拆解思维模式综述 | 引用级 | contexts/methodology/task_decomposition_thinking_survey_20260303_deep_research.md |
| human_decomposition_military_org_20260416_deep_research | 军事/组织拆解原则迁移 | 引用级 | contexts/methodology/human_decomposition_military_org_20260416_deep_research.md |
| human_decomposition_engineering_20260416_deep_research | 软工/管理史拆解方法 | 引用级 | contexts/methodology/human_decomposition_engineering_20260416_deep_research.md |
| long_context_methods_20260417_deep_research | RAG vs 长上下文决策方法 | 引用级 | contexts/methodology/long_context_methods_20260417_deep_research.md |
大上下文提取与信息收集 (6)
| slug | 一句话 | T1 分级 | 真源路径 |
|---|---|---|---|
| large_context_extraction_methodology_20260407_manual | 提取核心方法论（F1-F5 失败模式） | 实跑 | contexts/methodology/large_context_extraction_methodology_20260407_manual.md |
| large_context_final_methodology_20260408_manual | 信息瓶颈框架+验证原则 | 实跑 | contexts/methodology/large_context_final_methodology_20260408_manual.md |
| locate_read_two_phase_info_collection_20260530_manual → workflow_recall_first_locate | 定位/读取两阶段收集 | 实跑 memory_index | contexts/methodology/locate_read_two_phase_info_collection_20260530_manual.md |
| llm_context_inertia_and_task_allocation_20260429_manual | 注入 vs 基线 vs 注意力偏置机理 | 引用级 | contexts/methodology/llm_context_inertia_and_task_allocation_20260429_manual.md |
| text_extraction_methods_comprehensive_20260307_deep_research | 文本提取/摘要方法综述 | 引用级 | contexts/methodology/text_extraction_methods_comprehensive_20260307_deep_research.md |
| text_extraction_summarization_research_20260305_deep_research | NLP/embedding 方法综述 | 引用级 | contexts/methodology/text_extraction_summarization_research_20260305_deep_research.md |
编排与多 agent (8)
| slug | 一句话 | T1 分级 | 真源路径 |
|---|---|---|---|
| codex_sandbox_write_boundary_and_recovery_20260720_manual | Codex sandbox 写边界五类与恢复动作 | 引用级 memory_index | contexts/methodology/codex_sandbox_write_boundary_and_recovery_20260720_manual.md |
| orchestrator_prompt_patterns_20260417_manual | orchestrator prompt 模式库 | 引用级 | contexts/methodology/orchestrator_prompt_patterns_20260417_manual.md |
| interactive_coevolution_methodology_20260710_manual | 人机共演七步方法论 | 实跑 memory_index | contexts/methodology/interactive_coevolution_methodology_20260710_manual.md |
| agent_protocol_semantic_layer_excerpt_20260510_manual | agent 协议语义层框架 | 引用级 | contexts/methodology/agent_protocol_semantic_layer_excerpt_20260510_manual.md |
| agent_memory_survey_final_synthesis_20260226_deep_research | agent 记忆系统设计综合 | 引用级 | contexts/methodology/agent_memory_survey_final_synthesis_20260226_deep_research.md |
| agent_memory_survey_overview_20260226_deep_research | agent 记忆研究综述索引 | 引用级 | contexts/methodology/agent_memory_survey_overview_20260226_deep_research.md |
| subagent_failure_modes_and_entropy_control_survey_20260529 | sub-agent 失败模式与熵控制 | 引用级 memory_index | contexts/methodology/subagent_failure_modes_and_entropy_control_survey_20260529.md |
| subagent_conformity_reproduction_and_defense_20260530 | 抗从众实验方法与防御 | 实跑 memory_index | contexts/methodology/subagent_conformity_reproduction_and_defense_20260530.md |
provenance：contexts/methodology/INDEX.md adhoc_jobs/methodology_refactor_completion_audit_20260721/collect/T1_methodology_consumption_map.md