paput-mcp 2.7.0 → 2.8.0

package/docs/tools.md

@@ -16,8 +16,10 @@ Clients and assistants should follow these rules:
 - `paput_save_pending_candidate` requires explicit user approval because it creates a PaPut memo from a local pending candidate.
 - `paput_delete_skill_sheet_project` should be used only when the user clearly intends to remove a project.
 - `paput_update_skill_sheet_project_ai_summary` should be used only after the MCP client AI model has generated a project summary and the user intends to save it.
+- `paput_update_dashboard_analysis` should be used only after the MCP client AI model has generated dashboard analysis and the user intends to save it.
 - `paput_set_skill_sheet_skills` replaces the full skill list and should be used only when the complete desired final list is known.
 - `paput_discard_pending_candidate` removes a pending item from the save flow and should be confirmed when the candidate may still be useful.
+- `paput_delete_goal` should be used only when the user clearly intends to remove a goal.
 - Update and upsert tools should preserve existing data unless the user requested the change.
 
 ## Memo Tools
@@ -56,6 +58,35 @@ mode.
 | `paput_get_skill_sheet_project_summary_context` | Read-only         | Get project information and related memo bodies for MCP client-side summary generation.    |
 | `paput_update_skill_sheet_project_ai_summary`   | Destructive/write | Save an AI-generated project summary.                                                      |
 
+## Goal Tools
+
+| Tool                | Safety            | Use case                                                                            |
+| ------------------- | ----------------- | ----------------------------------------------------------------------------------- |
+| `paput_list_goals`  | Read-only         | List active and archived goals. Active goals are current analysis targets.          |
+| `paput_create_goal` | Write             | Create a goal for career, learning, portfolio, project, or other progress tracking. |
+| `paput_update_goal` | Destructive/write | Update a goal. The update request body includes `id`, and success returns no body.  |
+| `paput_delete_goal` | Destructive       | Delete a goal by ID. The API path uses `/goal/:id`.                                 |
+
+Archived goals are historical context. Active goals are normally used as the basis for
+dashboard analysis.
+
+## Dashboard Analysis Tools
+
+| Tool                                   | Safety            | Use case                                                                                        |
+| -------------------------------------- | ----------------- | ----------------------------------------------------------------------------------------------- |
+| `paput_get_dashboard_analysis`         | Read-only         | Get the saved dashboard analysis result.                                                        |
+| `paput_get_dashboard_analysis_context` | Read-only         | Get dashboard summary, goals, skill sheet, recent memos, notes, and categories for AI analysis. |
+| `paput_update_dashboard_analysis`      | Destructive/write | Save dashboard analysis generated by the MCP client AI model.                                   |
+
+`paput_get_dashboard_analysis_context` returns raw source data in
+`structuredContent` for MCP client-side analysis. The context asks the MCP client
+AI model to analyze current position, strengths, recently growing areas, thin
+areas, missing knowledge against goals, next knowledge to learn, and expressions
+suitable for a skill sheet or career history. It also instructs the client AI to
+write the final output in the user's language. The MCP server does not
+recalculate dashboard continuity from activities; it relies on summary values
+returned by the PaPut API.
+
 ## Knowledge Capture And Local Cache Tools
 
 These tools are local CLI mode only. They are not listed by the remote HTTP MCP

package/docs/usage-examples.md

@@ -104,3 +104,21 @@ Expected tool flow:
 3. `paput_get_note` to verify attached memo IDs
 
 Use case: create curated collections from existing PaPut memos.
+
+## 7. Generate And Save Dashboard Analysis
+
+Prompt:
+
+```text
+Analyze my PaPut dashboard and goals, then save the dashboard analysis.
+```
+
+Expected tool flow:
+
+1. `paput_get_dashboard_analysis_context`
+2. MCP client AI generates the analysis from `structuredContent`
+3. `paput_update_dashboard_analysis` after the user intends to save the result
+4. `paput_get_dashboard_analysis` to verify the saved result
+
+Use case: keep analysis generation in the MCP client AI while PaPut MCP only
+retrieves source data and saves the finished analysis.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "paput-mcp",
-  "version": "2.7.0",
+  "version": "2.8.0",
   "description": "PaPut MCP Server",
   "main": "dist/index.js",
   "type": "module",