dijk-skills 1.0.2 → 1.0.4

package/README.md

@@ -26,9 +26,9 @@ npx dijk-skills install [options]
 
 | Module | Skills |
 |--------|--------|
-| `common` | grilling, humanizer-zh, i-have-adhd |
-| `writing` | grill-my-insight, vibe-opinion-insight |
-| `tools` | darwin-skill, skill-creator |
+| `common` | caveman, grilling, handoff, humanizer-zh, i-have-adhd |
+| `writing` | grill-my-proposal, last30days, vibe-code-insight, vibe-opinion-insight |
+| `tools` | mcp-builder, skill-creator |
 | `coding` | *(empty, reserved for future skills)* |
 
 > Installing any module automatically includes `common`.
@@ -57,9 +57,10 @@ Skills are installed to `{dir}/skills/{skill-name}/`:
 ```
 ~/.agents/
 └── skills/
+    ├── caveman/
     ├── grilling/
+    ├── handoff/
     ├── humanizer-zh/
     ├── i-have-adhd/
-    ├── darwin-skill/
     └── ...
 ```

package/common/skills/handoff/SKILL.md

@@ -0,0 +1,16 @@
+---
+name: handoff
+description: Compact the current conversation into a handoff document for another agent to pick up.
+argument-hint: "What will the next session be used for?"
+disable-model-invocation: true
+---
+
+Write a handoff document summarising the current conversation so a fresh agent can continue the work. Save to the temporary directory of the user's OS - not the current workspace.
+
+Include a "suggested skills" section in the document, naming which skills the next agent should call the Skill tool for.
+
+Do not duplicate content already captured in other artifacts (specs, plans, ADRs, issues, commits, diffs). Reference them by path or URL instead.
+
+Redact any sensitive information, such as API keys, passwords, or personally identifiable information.
+
+If the user passed arguments, treat them as a description of what the next session will focus on and tailor the doc accordingly.

package/common/skills/handoff/agents/openai.yaml

@@ -0,0 +1,5 @@
+interface:
+  display_name: "Handoff"
+  short_description: "Compact a conversation into a handoff"
+policy:
+  allow_implicit_invocation: false

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "dijk-skills",
-  "version": "1.0.2",
+  "version": "1.0.4",
   "description": "CLI installer for Dijk skills collection",
   "type": "module",
   "bin": {

package/writing/skills/{grill-my-insight → grill-my-proposal}/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: grill-my-insight
+name: grill-my-proposal
 description: 把方案往死里问——三阶段递进追问（宏观结构→逐章深挖→交叉验证），从骨架到血肉逐层施压。一次只问一个问题，追问意图而非建议，不做决策。当用户想在动手前压力测试方案，或使用"挑战"、"拷问"、"作为蓝军"等触发词时启用。
 ---
 对方案进行三阶段递进追问，从骨架到血肉逐层施压，直到双方理解完全一致。

package/writing/skills/vibe-code-insight/SKILL.md

@@ -13,15 +13,14 @@ description: >
 
 ## Vibe 阶段
 
-在开始大规模研究之前，和用户简短对话，明确：
-- 读者是谁（架构师？决策者？学习者？）
-- 最想搞懂什么（系统怎么运作？和竞品的区别？值不值得押注？）
-- 有没有特别关注的维度（安全？性能？可扩展性？）
-- 架构视图偏好（4+1？C4？还是让分析者自选？）
+目标：最少交互，最大推断。
 
-这些问题决定了分析深度、视图组合和侧重点。不要假设，要问。
+1. **先读 prompt 推断** — 用户提到竞品就是竞品视角，说了"安全"就是安全维度
+2. **快速侦察项目** — README、目录结构、规模决定分析深度（500 行工具库不需要 4+1）
+3. **只补缺失维度** — 读者画像 / 核心关注 / 侧重维度 / 视图偏好，能推断的不问，最多问 2 个
+4. **展示推断，邀请纠正** — "判断你是 X 视角，重点 Y"，不要弹预设选项
 
-对话结束后，将结论写入 `insight/{slug}/brief.md`。后续每次触发 skill 时先读 brief.md 恢复意图，不需要重新对话。
+结论写入 `insight/{slug}/brief.md`，后续触发先读 brief 恢复意图。
 
 ## 交付物