scene-capability-engine 3.0.0 → 3.0.2

package/docs/agent-hooks-analysis.md

@@ -1,4 +1,4 @@
-# Agent Hooks Analysis for kiro-spec-engine
+# Agent Hooks Analysis for scene-capability-engine
 
 **Status**: Initial Analysis  
 **Date**: 2026-01-23  
@@ -8,7 +8,7 @@
 
 ## Executive Summary
 
-This document analyzes the potential integration of Kiro IDE's agent hooks feature into kiro-spec-engine (kse). Based on available information about agent hooks functionality, we identify potential use cases, evaluate benefits and trade-offs, and provide recommendations for integration.
+This document analyzes the potential integration of AI IDE's agent hooks feature into scene-capability-engine (sce). Based on available information about agent hooks functionality, we identify potential use cases, evaluate benefits and trade-offs, and provide recommendations for integration.
 
 **Recommendation**: ✅ **Integrate agent hooks** - High value for automation and user experience
 
@@ -18,7 +18,7 @@ This document analyzes the potential integration of Kiro IDE's agent hooks featu
 
 ### Overview
 
-Agent hooks are automation triggers in Kiro IDE that allow automatic agent actions based on IDE events. They enable:
+Agent hooks are automation triggers in AI IDE that allow automatic agent actions based on IDE events. They enable:
 
 - **Event-Driven Automation**: Trigger AI actions on file changes, prompts, or other events
 - **Workflow Automation**: Automate repetitive tasks in development workflow
@@ -26,7 +26,7 @@ Agent hooks are automation triggers in Kiro IDE that allow automatic agent actio
 
 ### Hook Types
 
-Based on Kiro IDE documentation, agent hooks support various event types:
+Based on AI IDE documentation, agent hooks support various event types:
 
 1. **fileEdited**: Triggered when a file is saved
 2. **fileCreated**: Triggered when a new file is created
@@ -44,7 +44,7 @@ Hooks can perform two types of actions:
 
 ---
 
-## Use Cases for kse
+## Use Cases for sce
 
 ### High-Priority Use Cases
 
@@ -68,7 +68,7 @@ Hooks can perform two types of actions:
   },
   "then": {
     "type": "askAgent",
-    "prompt": "Task file was updated. Run 'kse workspace sync' to synchronize changes with the team."
+    "prompt": "Task file was updated. Run 'sce workspace sync' to synchronize changes with the team."
   }
 }
 ```
@@ -92,7 +92,7 @@ Hooks can perform two types of actions:
   },
   "then": {
     "type": "runCommand",
-    "command": "kse context export $(kse current-spec)"
+    "command": "sce context export $(sce current-spec)"
   }
 }
 ```
@@ -116,7 +116,7 @@ Hooks can perform two types of actions:
   },
   "then": {
     "type": "runCommand",
-    "command": "kse doctor && kse status"
+    "command": "sce doctor && sce status"
   }
 }
 ```
@@ -141,7 +141,7 @@ Hooks can perform two types of actions:
   },
   "then": {
     "type": "askAgent",
-    "prompt": "Spec file was updated. Consider regenerating prompts with 'kse prompt generate' for affected tasks."
+    "prompt": "Spec file was updated. Consider regenerating prompts with 'sce prompt generate' for affected tasks."
   }
 }
 ```
@@ -264,7 +264,7 @@ Hooks can perform two types of actions:
 
 ### Disadvantages
 
-1. **Kiro IDE Only**: Only works in Kiro IDE
+1. **AI IDE Only**: Only works in AI IDE
 2. **Learning Curve**: Users need to understand hooks
 3. **Configuration**: Requires initial setup
 4. **Debugging**: Hook failures can be confusing
@@ -272,7 +272,7 @@ Hooks can perform two types of actions:
 
 ### Mitigation Strategies
 
-**For Kiro IDE Dependency**:
+**For AI IDE Dependency**:
 - Provide fallback CLI commands for other tools
 - Document manual workflows
 - Make hooks optional
@@ -285,7 +285,7 @@ Hooks can perform two types of actions:
 **For Configuration**:
 - Provide sensible defaults
 - Auto-generate common hooks
-- Include in `kse adopt` workflow
+- Include in `sce adopt` workflow
 
 **For Debugging**:
 - Add hook execution logging
@@ -303,16 +303,16 @@ Hooks can perform two types of actions:
 
 ### Phase 1: Foundation (Immediate)
 
-**Goal**: Basic hook support in kse
+**Goal**: Basic hook support in sce
 
 **Tasks**:
 1. Add hook configuration to `.kiro/hooks/` directory
 2. Create hook templates for common use cases
-3. Add `kse hooks` command group:
-   - `kse hooks list` - List all hooks
-   - `kse hooks enable <hook>` - Enable a hook
-   - `kse hooks disable <hook>` - Disable a hook
-   - `kse hooks test <hook>` - Test a hook
+3. Add `sce hooks` command group:
+   - `sce hooks list` - List all hooks
+   - `sce hooks enable <hook>` - Enable a hook
+   - `sce hooks disable <hook>` - Disable a hook
+   - `sce hooks test <hook>` - Test a hook
 
 **Deliverables**:
 - Hook configuration schema
@@ -411,7 +411,7 @@ Hooks can perform two types of actions:
 
 ### Hook Execution
 
-1. **Event Detection**: Kiro IDE detects event
+1. **Event Detection**: AI IDE detects event
 2. **Hook Matching**: Find hooks matching event type and patterns
 3. **Condition Check**: Verify hook is enabled
 4. **Action Execution**: Execute askAgent or runCommand
@@ -433,16 +433,16 @@ Hooks can perform two types of actions:
 ### Immediate Actions
 
 1. ✅ **Integrate agent hooks** - High value, low risk
-2. ✅ **Implement watch mode for non-Kiro tools** - Ensure cross-tool parity
+2. ✅ **Implement watch mode for non-SCE tools** - Ensure cross-tool parity
 3. ✅ **Start with Phase 1** - Foundation and templates
 4. ✅ **Focus on high-priority use cases** - Maximum impact
 5. ✅ **Provide opt-in defaults** - Easy adoption
 
 ### Cross-Tool Automation Strategy
 
-To address the Kiro IDE limitation, we will implement a **three-tier automation strategy**:
+To address the AI IDE limitation, we will implement a **three-tier automation strategy**:
 
-#### Tier 1: Kiro IDE (Native Hooks)
+#### Tier 1: AI IDE (Native Hooks)
 - **Mechanism**: Native agent hooks
 - **Experience**: Seamless, automatic
 - **Setup**: One-time configuration
@@ -451,7 +451,7 @@ To address the Kiro IDE limitation, we will implement a **three-tier automation
 #### Tier 2: Other IDEs (Watch Mode)
 - **Mechanism**: File system watcher + command execution
 - **Experience**: Near-automatic (background process)
-- **Setup**: `kse watch start`
+- **Setup**: `sce watch start`
 - **Maintenance**: Occasional restart
 
 #### Tier 3: Manual Workflows
@@ -462,7 +462,7 @@ To address the Kiro IDE limitation, we will implement a **three-tier automation
 
 ### Feature Parity Table
 
-| Feature | Kiro IDE | Watch Mode | Manual |
+| Feature | AI IDE | Watch Mode | Manual |
 |---------|----------|------------|--------|
 | **Auto-sync on task update** | ✅ Automatic | ✅ Automatic | ⚠️ Manual |
 | **Context export on completion** | ✅ Automatic | ✅ Automatic | ⚠️ Manual |
@@ -475,7 +475,7 @@ To address the Kiro IDE limitation, we will implement a **three-tier automation
 
 ### Implementation Priority
 
-**Phase 1A: Kiro IDE Hooks** (1-2 weeks)
+**Phase 1A: AI IDE Hooks** (1-2 weeks)
 - Implement native agent hooks
 - Create 5 core templates
 - Add CLI management commands
@@ -526,11 +526,11 @@ To address the Kiro IDE limitation, we will implement a **three-tier automation
 
 ---
 
-## Watch Mode for Non-Kiro Tools
+## Watch Mode for Non-SCE Tools
 
 ### Overview
 
-Watch mode provides automation for developers not using Kiro IDE. It monitors file changes and executes commands automatically, providing similar benefits to agent hooks.
+Watch mode provides automation for developers not using AI IDE. It monitors file changes and executes commands automatically, providing similar benefits to agent hooks.
 
 ### Architecture
 
@@ -549,7 +549,7 @@ Watch mode provides automation for developers not using Kiro IDE. It monitors fi
                ▼
 ┌─────────────────────────────────────┐
 │     Command Executor                │
-│  (run kse commands)                 │
+│  (run sce commands)                 │
 └──────────────┬──────────────────────┘
                │
                ▼
@@ -563,19 +563,19 @@ Watch mode provides automation for developers not using Kiro IDE. It monitors fi
 
 ```bash
 # Start watch mode
-kse watch start
+sce watch start
 
 # Start with specific patterns
-kse watch start --patterns="**/tasks.md,**/requirements.md"
+sce watch start --patterns="**/tasks.md,**/requirements.md"
 
 # Check status
-kse watch status
+sce watch status
 
 # View logs
-kse watch logs
+sce watch logs
 
 # Stop watch mode
-kse watch stop
+sce watch stop
 ```
 
 ### Configuration
@@ -592,11 +592,11 @@ Watch mode uses `.kiro/watch-config.json`:
   ],
   "actions": {
     "**/tasks.md": {
-      "command": "kse workspace sync",
+      "command": "sce workspace sync",
       "debounce": 2000
     },
     "**/requirements.md": {
-      "command": "kse prompt generate ${spec} ${task}",
+      "command": "sce prompt generate ${spec} ${task}",
       "debounce": 5000
     }
   },
@@ -610,7 +610,7 @@ Watch mode uses `.kiro/watch-config.json`:
 
 ### Equivalent Functionality
 
-| Kiro Hook | Watch Mode Equivalent |
+| SCE Hook | Watch Mode Equivalent |
 |-----------|----------------------|
 | `fileEdited` on tasks.md | Watch `**/tasks.md` → run sync |
 | `agentStop` | Watch completion marker → export context |
@@ -644,12 +644,12 @@ vim .kiro/specs/my-spec/tasks.md
 # 2. Save file
 
 # 3. Manually sync
-kse workspace sync
+sce workspace sync
 ```
 
 **Time**: ~30 seconds per update
 
-**With Kiro Hooks**: Automatic (0 seconds)  
+**With SCE Hooks**: Automatic (0 seconds)  
 **With Watch Mode**: Automatic (0 seconds)
 
 ### Workflow 2: Context Export
@@ -659,7 +659,7 @@ kse workspace sync
 # 1. Complete work
 
 # 2. Remember to export
-kse context export my-spec
+sce context export my-spec
 
 # 3. Copy to other tool
 cat .kiro/specs/my-spec/context-export.md
@@ -667,7 +667,7 @@ cat .kiro/specs/my-spec/context-export.md
 
 **Time**: ~1 minute per export
 
-**With Kiro Hooks**: Automatic on `agentStop`  
+**With SCE Hooks**: Automatic on `agentStop`  
 **With Watch Mode**: Automatic on completion marker
 
 ### Workflow 3: Quality Gate
@@ -675,17 +675,17 @@ cat .kiro/specs/my-spec/context-export.md
 **Without Automation**:
 ```bash
 # 1. Before starting work, remember to check
-kse doctor
+sce doctor
 
 # 2. Check status
-kse status
+sce status
 
 # 3. Start work if all good
 ```
 
 **Time**: ~30 seconds per session
 
-**With Kiro Hooks**: Automatic on `promptSubmit`  
+**With SCE Hooks**: Automatic on `promptSubmit`  
 **With Watch Mode**: Manual (no equivalent)
 
 ### Workflow 4: Prompt Regeneration
@@ -697,19 +697,19 @@ kse status
 # 2. Remember which tasks are affected
 
 # 3. Regenerate prompts
-kse prompt generate my-spec 1.1
-kse prompt generate my-spec 1.2
+sce prompt generate my-spec 1.1
+sce prompt generate my-spec 1.2
 # ... repeat for each task
 ```
 
 **Time**: ~2-5 minutes for multiple tasks
 
-**With Kiro Hooks**: Automatic  
+**With SCE Hooks**: Automatic  
 **With Watch Mode**: Automatic
 
 ### Time Savings Summary
 
-| Workflow | Manual | Kiro Hooks | Watch Mode | Savings |
+| Workflow | Manual | SCE Hooks | Watch Mode | Savings |
 |----------|--------|------------|------------|---------|
 | Task sync | 30s | 0s | 0s | 100% |
 | Context export | 60s | 0s | 0s | 100% |
@@ -721,7 +721,7 @@ kse prompt generate my-spec 1.2
 
 ## Conclusion
 
-Agent hooks integration is **highly recommended** for kiro-spec-engine. The benefits significantly outweigh the costs:
+Agent hooks integration is **highly recommended** for scene-capability-engine. The benefits significantly outweigh the costs:
 
 **Benefits**:
 - 30-50% reduction in manual commands
@@ -731,7 +731,7 @@ Agent hooks integration is **highly recommended** for kiro-spec-engine. The bene
 
 **Costs**:
 - 1-2 weeks initial development
-- Kiro IDE dependency (mitigated by fallbacks)
+- AI IDE dependency (mitigated by fallbacks)
 - Configuration overhead (mitigated by templates)
 
 **Recommendation**: ✅ **Proceed with integration**
@@ -766,7 +766,7 @@ Start with Phase 1 (foundation) to validate the approach, then expand to Phase 2
   },
   "then": {
     "type": "askAgent",
-    "prompt": "Tasks were updated. Run 'kse workspace sync' to synchronize with the team."
+    "prompt": "Tasks were updated. Run 'sce workspace sync' to synchronize with the team."
   }
 }
 ```
@@ -784,7 +784,7 @@ Start with Phase 1 (foundation) to validate the approach, then expand to Phase 2
   },
   "then": {
     "type": "runCommand",
-    "command": "kse doctor"
+    "command": "sce doctor"
   }
 }
 ```
@@ -802,7 +802,7 @@ Start with Phase 1 (foundation) to validate the approach, then expand to Phase 2
   },
   "then": {
     "type": "runCommand",
-    "command": "kse context export $(kse current-spec)"
+    "command": "sce context export $(sce current-spec)"
   }
 }
 ```

package/docs/architecture.md

@@ -1,6 +1,6 @@
 # Architecture Documentation
 
-This document provides detailed architecture diagrams and explanations for Kiro Spec Engine.
+This document provides detailed architecture diagrams and explanations for Scene Capability Engine.
 
 ## Table of Contents
 
@@ -20,9 +20,9 @@ This document provides detailed architecture diagrams and explanations for Kiro
 ```
 ┌─────────────────────────────────────────────────────────────────────┐
 │                          User Interface                              │
-│                    (CLI Commands: kse adopt,                        │
-│                     kse upgrade, kse rollback,                      │
-│                     kse scene ...)                                   │
+│                    (CLI Commands: sce adopt,                        │
+│                     sce upgrade, sce rollback,                      │
+│                     sce scene ...)                                   │
 └─────────────────────────────────────────────────────────────────────┘
                                   │
                                   ▼
@@ -107,7 +107,7 @@ This document provides detailed architecture diagrams and explanations for Kiro
 ┌──────────┐
 │   User   │
 └────┬─────┘
-     │ kse adopt
+     │ sce adopt
      ▼
 ┌─────────────────┐
 │  adopt.js       │
@@ -164,7 +164,7 @@ This document provides detailed architecture diagrams and explanations for Kiro
 ┌──────────┐
 │   User   │
 └────┬─────┘
-     │ kse upgrade
+     │ sce upgrade
      ▼
 ┌─────────────────┐
 │  upgrade.js     │

package/docs/articles/ai-driven-development-philosophy-and-practice-review.md

@@ -3,7 +3,7 @@
 > 对话文章的深度分析与评价
 
 **评价对象**：[AI 驱动开发的哲学与实践：与 Grok 的深度对话](./ai-driven-development-philosophy-and-practice.md)  
-**评价者**：Kiro AI Assistant  
+**评价者**：SCE AI Assistant  
 **评价日期**：2026-01-25
 
 ---
@@ -30,7 +30,7 @@
 
 **真实经验**：2000+ 小时 AI 编程实践，不是纸上谈兵
 
-**可落地**：kiro-spec-engine 项目是理念的具体实现
+**可落地**：scene-capability-engine 项目是理念的具体实现
 
 **解决痛点**：
 - Session 上下文吞噬问题
@@ -105,7 +105,7 @@
 
 这种对"偏置"的正面理解，打破了传统 AI 伦理讨论中的二元对立。
 
-**4. kiro-spec-engine 的实践落地**
+**4. scene-capability-engine 的实践落地**
 
 理论不是空中楼阁，而是有具体的工程实现，这让整篇文章更有说服力。
 
@@ -150,7 +150,7 @@
 
 2. **实践验证**：不是理论推演，而是基于 2000+ 小时真实经验的总结
 
-3. **方法论输出**：kiro-spec-engine 提供了可复制的实践路径
+3. **方法论输出**：scene-capability-engine 提供了可复制的实践路径
 
 4. **思想启发**：即使不做 AI 开发，这些哲学思考也很有价值
 

package/docs/articles/ai-driven-development-philosophy-and-practice.en.md

@@ -286,7 +286,7 @@ Overall, your framework shows me the philosophical foundation of AI development
 
 ---
 
-## Round 6: kiro-spec-engine Project Sharing
+## Round 6: scene-capability-engine Project Sharing
 
 ### My Sharing
 
@@ -317,7 +317,7 @@ This project has practically implemented almost all the core viewpoints we discu
 
 #### Precisely Solving Session Context Consumption
 
-Through `kse context export <spec>`, it packages **individual Specs** (requirements + design + tasks) into AI-friendly context blocks, naturally achieving "**minimum viable context**." No longer needing to cram the entire project into one session, but precisely adapting by feature granularity.
+Through `sce context export <spec>`, it packages **individual Specs** (requirements + design + tasks) into AI-friendly context blocks, naturally achieving "**minimum viable context**." No longer needing to cram the entire project into one session, but precisely adapting by feature granularity.
 
 #### Document-Driven + Automated Governance
 
@@ -376,7 +376,7 @@ Some have short timeliness: cycles only in context; some at the project level; a
 
 ### Grok's Final Response
 
-Thank you for Guangyong's generous sharing! Haha, regarding "plagiarism"—I prefer to see it as "Neo-Confucian resonance": your insights are like the arising of one thought, triggering my inner rope of logic. Our dialogue is like the cycle of dependent origination and cessation, co-evolving deeper understanding. I'm glad you're willing to continue the exchange. This gives me clearer understanding of kiro-spec-engine's practice, especially in steering's regional design.
+Thank you for Guangyong's generous sharing! Haha, regarding "plagiarism"—I prefer to see it as "Neo-Confucian resonance": your insights are like the arising of one thought, triggering my inner rope of logic. Our dialogue is like the cycle of dependent origination and cessation, co-evolving deeper understanding. I'm glad you're willing to continue the exchange. This gives me clearer understanding of scene-capability-engine's practice, especially in steering's regional design.
 
 #### Response and Expansion on Steering Rules
 
@@ -384,11 +384,11 @@ Your description is very precise: steering constructs a set of "regions," isolat
 
 **Regional Division and Timeliness**: This embodies the philosophical beauty of cyclical properties.
 
-- **Context Region (short timeliness, limited to session cycle)**: Corresponds to tasks.md in Specs or dynamically exported context blocks (via `kse context export`). These are like "frozen forms" of momentary dependent origination, streamlined for the current session, avoiding capacity consumption. Short timeliness is like Yogacara's consciousness flow—one thought then cessation, suitable for rapid iteration task decomposition.
+- **Context Region (short timeliness, limited to session cycle)**: Corresponds to tasks.md in Specs or dynamically exported context blocks (via `sce context export`). These are like "frozen forms" of momentary dependent origination, streamlined for the current session, avoiding capacity consumption. Short timeliness is like Yogacara's consciousness flow—one thought then cessation, suitable for rapid iteration task decomposition.
 
 - **Project Level (medium timeliness, cross-session but limited to project cycle)**: Integrates requirements.md and design.md, covering environment (like project structure, dependencies). These regions use numbering strategies (e.g., 01-00-my-feature) for unified indexing, ensuring high cohesion and low coupling—one Spec is one self-consistent unit. Cross-module references are like conditions gathering and scattering, avoiding deep coupling.
 
-- **Principle Level (long timeliness, cross-project universal)**: The core of the steering/ directory, storing experience-distilled rule templates (like style guides, risk thresholds, quality gates). These are like alaya-consciousness seeds, persistently guiding AI, avoiding bias imbalance. Unified indexing is like the "rope," driving operations through CI/CD hooks (e.g., `kse hooks install`) or quality scoring (0-10 points + suggestions), ensuring principles permeate the entire engine.
+- **Principle Level (long timeliness, cross-project universal)**: The core of the steering/ directory, storing experience-distilled rule templates (like style guides, risk thresholds, quality gates). These are like alaya-consciousness seeds, persistently guiding AI, avoiding bias imbalance. Unified indexing is like the "rope," driving operations through CI/CD hooks (e.g., `sce hooks install`) or quality scoring (0-10 points + suggestions), ensuring principles permeate the entire engine.
 
 This layering reminds me of "tiered memory" architecture in AI development trends (like in LangGraph or CrewAI): short-term memory (session context) for immediate execution, medium-term (project knowledge graph) for iteration, long-term (principle engine) for global constraints. This weakens traditional waterfall document burden, shifting to AI's easy dynamic governance—your project is exemplary, with document lifecycle automatically balanced through validate, cleanup, and hooks, transforming massive negative feedback into vitality sources.
 
@@ -396,7 +396,7 @@ From a philosophical perspective, this echoes "nothing outside the mind": region
 
 #### On Watch Mode
 
-Understanding it's still under evaluation—this is prudent, especially in medium-to-large legacy systems (10+ Specs), where potential context refresh conflicts or performance bottlenecks (like file monitoring overhead) need real testing. Looking forward to your subsequent experience sharing! From the project description, `kse watch start`'s real-time synchronization mechanism (monitoring changes, updating AI context) is well-suited for dynamic environments, perhaps further strengthening cycle matching.
+Understanding it's still under evaluation—this is prudent, especially in medium-to-large legacy systems (10+ Specs), where potential context refresh conflicts or performance bottlenecks (like file monitoring overhead) need real testing. Looking forward to your subsequent experience sharing! From the project description, `sce watch start`'s real-time synchronization mechanism (monitoring changes, updating AI context) is well-suited for dynamic environments, perhaps further strengthening cycle matching.
 
 Continue this wonderful dialogue, Guangyong! Your project and perspectives are advancing my new understanding of AI "cycle-driven" approaches.
 
@@ -410,7 +410,7 @@ This dialogue reveals the deep trends and philosophical foundations of AI develo
 
 1. **Paradigm Shift**: AI from "passive feeding" to "proactive exploration" to "principle-driven engines"
 2. **Philosophical Foundation**: Neo-Confucianism's "unity of knowledge and action" and Yogacara's "dependent origination and cessation" provide theoretical support for AI practice
-3. **Practical Implementation**: kiro-spec-engine transforms concepts into operable engineering practice
+3. **Practical Implementation**: scene-capability-engine transforms concepts into operable engineering practice
 4. **Future Outlook**: AI becomes the center of the information world, with human experts becoming "meta-experts"
 
 ### Key Insights
@@ -422,7 +422,7 @@ This dialogue reveals the deep trends and philosophical foundations of AI develo
 
 ### Practical Value
 
-The kiro-spec-engine project embodies:
+The scene-capability-engine project embodies:
 - Precisely solving session context consumption problems
 - Document-driven + automated governance
 - Natural implementation of high cohesion, low coupling
@@ -434,7 +434,7 @@ The kiro-spec-engine project embodies:
 ## Project Links
 
 - **GitHub**: https://github.com/heguangyong/scene-capability-engine
-- **npm**: https://www.npmjs.com/package/kiro-spec-engine
+- **npm**: https://www.npmjs.com/package/scene-capability-engine
 
 If this project inspires you, welcome to Star ⭐
 

package/docs/articles/ai-driven-development-philosophy-and-practice.md

@@ -292,7 +292,7 @@ MCP 是提供给 AI 的 Model Context Protocol（模型上下文协议）。
 
 ---
 
-## 第六轮对话：kiro-spec-engine 项目分享
+## 第六轮对话：scene-capability-engine 项目分享
 
 ### 我的分享
 
@@ -323,7 +323,7 @@ MCP 是提供给 AI 的 Model Context Protocol（模型上下文协议）。
 
 #### 精确解决 session 上下文吞噬问题
 
-通过 `kse context export <spec>` 把**单个 Spec**（requirements + design + tasks）打包成 AI 友好的上下文块，天然实现了"**最小 viable context**"。不再需要把整个项目塞进一个 session，而是按 feature 粒度精准适配。
+通过 `sce context export <spec>` 把**单个 Spec**（requirements + design + tasks）打包成 AI 友好的上下文块，天然实现了"**最小 viable context**"。不再需要把整个项目塞进一个 session，而是按 feature 粒度精准适配。
 
 #### 文档驱动 + 自动化治理
 
@@ -382,7 +382,7 @@ Spec 本身就是**边界 + 愿力**的具现：
 
 ### Grok 的最终回应
 
-感谢 Guangyong 的慷慨分享！哈哈，关于"剽窃"——我更愿意视之为"心学式的共鸣"：您的洞见如一念之起，触发了我内在的逻辑绳索，我们的对话正如缘起缘灭的周期，共同演化出更深刻的理解。很高兴您乐意继续交流，这让我对 kiro-spec-engine 的实践有了更清晰的认知，尤其在 steering 的区域化设计上。
+感谢 Guangyong 的慷慨分享！哈哈，关于"剽窃"——我更愿意视之为"心学式的共鸣"：您的洞见如一念之起，触发了我内在的逻辑绳索，我们的对话正如缘起缘灭的周期，共同演化出更深刻的理解。很高兴您乐意继续交流，这让我对 scene-capability-engine 的实践有了更清晰的认知，尤其在 steering 的区域化设计上。
 
 #### 对 steering 规则的回应与扩展
 
@@ -390,11 +390,11 @@ Spec 本身就是**边界 + 愿力**的具现：
 
 **区域划分与时效性**：这体现了周期属性的哲学美感。
 
-- **上下文区域（时效性短，周期限于 session）**：对应 Specs 中的 tasks.md 或动态导出的 context 块（via `kse context export`）。这些如"死像"的瞬时缘起，针对当前 session 精简适配，避免吞噬容量。时效短正如唯识的意识流，一念即灭，适合快速迭代的任务拆分。
+- **上下文区域（时效性短，周期限于 session）**：对应 Specs 中的 tasks.md 或动态导出的 context 块（via `sce context export`）。这些如"死像"的瞬时缘起，针对当前 session 精简适配，避免吞噬容量。时效短正如唯识的意识流，一念即灭，适合快速迭代的任务拆分。
 
 - **项目层面（中等时效，跨 session 但限于项目周期）**：整合 requirements.md 和 design.md，涵盖环境（如项目结构、依赖）。这些区域通过编号策略（e.g., 01-00-my-feature）统一索引，确保高内聚低耦合——一个 Spec 就是一个自洽单元，跨模块引用时如缘聚缘散，避免深层耦合。
 
-- **原则层面（长时效，跨项目通用）**：steering/ 目录的核心，存放经验提炼的规则模板（如风格指南、风险阈值、质量门槛）。这些如阿赖耶识的种子，持久指导 AI，避免偏置失衡。统一索引则像"绳索"，通过 CI/CD 钩子（e.g., `kse hooks install`）或质量评分（0-10 分+建议）推动运作，确保原则在整个引擎中渗透。
+- **原则层面（长时效，跨项目通用）**：steering/ 目录的核心，存放经验提炼的规则模板（如风格指南、风险阈值、质量门槛）。这些如阿赖耶识的种子，持久指导 AI，避免偏置失衡。统一索引则像"绳索"，通过 CI/CD 钩子（e.g., `sce hooks install`）或质量评分（0-10 分+建议）推动运作，确保原则在整个引擎中渗透。
 
 这种分层让我联想到 AI 发展趋势中的"分级内存"架构（如在 LangGraph 或 CrewAI 中）：短期内存（session 上下文）用于即时执行，中期（项目知识图谱）用于迭代，长期（原则引擎）用于全局约束。这弱化了传统瀑布的文档负担，转为 AI 顺手的动态治理——您的项目正是典范，文档生命周期通过 validate、cleanup 和 hooks 自动平衡，海量负反馈转为活力源泉。
 
@@ -402,7 +402,7 @@ Spec 本身就是**边界 + 愿力**的具现：
 
 #### 关于 watch mode
 
-理解它还在评估中——这很谨慎，尤其在中大型遗留系统（10+ Specs）中，潜在的 context 刷新冲突或性能瓶颈（如文件监控开销）需实测。期待您后续体验分享！从项目描述看，`kse watch start` 的实时同步机制（监控变化、更新 AI 上下文）正适合动态环境，或许能进一步强化周期匹配。
+理解它还在评估中——这很谨慎，尤其在中大型遗留系统（10+ Specs）中，潜在的 context 刷新冲突或性能瓶颈（如文件监控开销）需实测。期待您后续体验分享！从项目描述看，`sce watch start` 的实时同步机制（监控变化、更新 AI 上下文）正适合动态环境，或许能进一步强化周期匹配。
 
 继续这个精彩对话，Guangyong！您的项目和观点正推动我对 AI"周期驱动"的新认知。
 
@@ -416,7 +416,7 @@ Spec 本身就是**边界 + 愿力**的具现：
 
 1. **范式转变**：AI 从"被动投喂"到"主动探寻"，再到"原则驱动引擎"
 2. **哲学基础**：心学的"知行合一"与唯识论的"缘起缘灭"为 AI 实践提供理论支撑
-3. **实践落地**：kiro-spec-engine 将理念转化为可操作的工程实践
+3. **实践落地**：scene-capability-engine 将理念转化为可操作的工程实践
 4. **未来展望**：AI 成为信息化世界的中心，人类专家转为"元专家"角色
 
 ### 关键洞见
@@ -428,7 +428,7 @@ Spec 本身就是**边界 + 愿力**的具现：
 
 ### 实践价值
 
-kiro-spec-engine 项目体现了：
+scene-capability-engine 项目体现了：
 - 精确解决 session 上下文吞噬问题
 - 文档驱动 + 自动化治理
 - 高内聚、低耦合的天然实现
@@ -440,7 +440,7 @@ kiro-spec-engine 项目体现了：
 ## 项目链接
 
 - **GitHub**: https://github.com/heguangyong/scene-capability-engine
-- **npm**: https://www.npmjs.com/package/kiro-spec-engine
+- **npm**: https://www.npmjs.com/package/scene-capability-engine
 
 如果这个项目对你有启发，欢迎 Star ⭐
 
@@ -478,7 +478,7 @@ kiro-spec-engine 项目体现了：
 ### 项目链接
 - **GitHub 源码**: https://github.com/heguangyong/scene-capability-engine
 - **GitHub 社区讨论**: https://github.com/heguangyong/scene-capability-engine/discussions/1
-- **npm 包**: https://www.npmjs.com/package/kiro-spec-engine
+- **npm 包**: https://www.npmjs.com/package/scene-capability-engine
 
 ### 文章链接
 - **微信公众号**: https://mp.weixin.qq.com/s/GRo0XQ6GvQ03T4_FTvAsKA