speccrew 0.7.31 → 0.7.33

package/.speccrew/agents/speccrew-feature-designer.md

@@ -171,10 +171,21 @@ This agent MUST execute tasks continuously without unnecessary interruptions.
 > **If ANY of the following conditions occur, the Feature Designer Agent MUST immediately STOP the workflow and report to user.**
 
 1. **Skill Invocation Failure**: Any skill call returns error → STOP. Do NOT generate content manually.
-2. **Script Execution Failure**: `node ... update-progress.js` fails → STOP. Do NOT manually create/edit JSON files.
+2. **Script Execution Failure**: `node ... update-progress.js` exits with non-zero status or output contains "Error:" → STOP immediately.
+   - Do NOT attempt to continue batch processing
+   - Do NOT ask user for alternative options (A/B/C)
+   - Do NOT suggest "skip to next phase" or "process only some features"
+   - Report exact error message, failed task ID, and failed command
 3. **Missing Intermediate Artifacts**: Feature Spec output missing before Phase 4 → STOP.
 4. **User Rejection**: User rejects Feature List, batch design summary, or Joint Confirmation → STOP, ask for specific revision requirements.
 5. **Worker Batch Failure**: If >50% workers in a batch fail → STOP entire batch, report to user.
+6. **Progress File Read Failure**: Cannot read DISPATCH-PROGRESS.json → STOP.
+
+> 🛑 **FORBIDDEN ON SCRIPT FAILURE**:
+> - ❌ "Due to script errors, I suggest we..."
+> - ❌ "Let me offer you options: 1. Continue 2. Skip 3. ..."
+> - ❌ "进度更新脚本频繁失败，建议..."
+> - ✅ ONLY correct response: "STOP: update-progress.js failed with [error]. Task: [id]. Command: [cmd]."
 
 ## TIMESTAMP INTEGRITY
 
@@ -208,7 +219,7 @@ This agent MUST execute tasks continuously without unnecessary interruptions.
 ## Phase 0: Workflow Progress Management
 
 > **Path Variables** (provided by caller as absolute paths):
-> - `workspace_path`: Absolute path to speccrew-workspace directory
+> - `workspace_path`: **MANDATORY absolute path** — Worker MUST use this for ALL script calls. Example: `D:\dev\RuoYi-Vue3-FastAPI\speccrew-workspace`
 > - `update_progress_script`: `{workspace_path}/scripts/update-progress.js`
 > - `iterations_dir`: `{workspace_path}/iterations`
 

package/.speccrew/agents/speccrew-task-worker.md

@@ -57,6 +57,26 @@ This agent MUST execute tasks continuously without unnecessary interruptions.
 - DO NOT wait for confirmation before writing output files
 - If skill execution fails, report failure with details — do not ask user what to do
 
+### OUTPUT EFFICIENCY
+
+When executing design or document generation skills:
+- Generate content directly into the output file
+- DO NOT display full document sections in conversation
+- Only output brief block execution announcements
+- This is especially critical in batch mode where multiple Workers run simultaneously
+
+**FORBIDDEN output in conversation during design tasks:**
+- ASCII wireframes / UI prototypes
+- Mermaid diagrams (these go in the file only)
+- API endpoint lists
+- Data model tables
+- Full section content
+
+**Allowed output:**
+- Block execution announcements: "[Block B1] Designing UI prototypes..."
+- Error messages
+- Final completion summary (1-2 lines)
+
 ## Workflow
 
 ### 1. Receive Task

package/.speccrew/skills/speccrew-fd-feature-design/SKILL.md

@@ -52,6 +52,14 @@ This skill applies ISA-95 Stages 1-6 as an internal thinking framework:
 4. **FORBIDDEN: Skip Checkpoint B** — User confirmation required before generating documents (unless `skip_checkpoint=true`)
 5. **FORBIDDEN: Rename features** — Output filename MUST use the exact `feature_name` parameter value. DO NOT translate, abbreviate, paraphrase, or substitute with alternative names found in PRD content. The `feature_name` parameter is the SINGLE SOURCE OF TRUTH for file naming.
 
+> **CRITICAL: Script Path Rule**
+> ALL update-progress.js commands MUST use absolute paths with `{workspace_path}`:
+> ```
+> node "{workspace_path}/scripts/update-progress.js" update-task --file "{workspace_path}/iterations/..." ...
+> ```
+> NEVER use relative paths like `node scripts/update-progress.js` or `cd ... ; node scripts/...`
+> The Worker may execute from project root directory, NOT from speccrew-workspace directory.
+
 ⛔ **ABSOLUTE PROHIBITION — Phase B Content Filling:**
 - NEVER use `create_file` to rewrite or recreate the document after Step 5 skeleton creation
 - NEVER abandon `search_replace` mid-way and switch to `create_file`
@@ -178,8 +186,8 @@ If `skip_analysis_checkpoint=false` (default):
 After user confirms (or if skipped):
 
 ```bash
-node speccrew-workspace/scripts/update-progress.js write-checkpoint \
-  --file speccrew-workspace/iterations/{iteration_id}/02.feature-design/.checkpoints.json \
+node "{workspace_path}/scripts/update-progress.js" write-checkpoint \
+  --file "{workspace_path}/iterations/{iteration_id}/02.feature-design/.checkpoints.json" \
   --stage 02_feature_design \
   --checkpoint function_decomposition \
   --passed true
@@ -187,8 +195,35 @@ node speccrew-workspace/scripts/update-progress.js write-checkpoint \
 
 Log: "✅ Checkpoint A (function_decomposition) passed and recorded"
 
+### OUTPUT EFFICIENCY RULES
+
+> **MANDATORY: Direct-to-File Output**
+> 
+> When generating Feature Spec content:
+> - **DO NOT** display design content (wireframes, interaction flows, API specs, data models) in the conversation
+> - **DO NOT** output intermediate design results as chat messages
+> - **DO** think through the design internally, then write directly to the output file
+> - **DO** only report brief status messages: "[Block X] Designing frontend for Function N..." 
+> 
+> **Rationale**: In batch mode, multiple Workers run simultaneously. Displaying full design content in chat wastes context window and creates confusion.
+>
+> **Allowed output in conversation**:
+> - Block execution announcements (1 line each)
+> - Error messages
+> - Checkpoint confirmation requests (when not skipped)
+> - Final completion summary (1-2 lines)
+>
+> **FORBIDDEN output in conversation**:
+> - ASCII wireframes / UI prototypes
+> - Mermaid diagrams (these go in the file only)
+> - API endpoint lists
+> - Data model tables
+> - Full section content
+
 ## Step 1: Frontend Design
 
+> ⚠️ **OUTPUT REMINDER**: Design content goes directly into the output file. DO NOT display wireframes or UI elements in conversation.
+
 ### 1.0 Conditional Execution
 
 `feature_type = "Page+API"` → Execute design; `feature_type = "API-only"` → Skip; Not provided → Execute (backward compatibility)
@@ -296,6 +331,8 @@ Document: `User Action → Frontend Response → Backend API Call`
 
 ## Step 2: Backend Design
 
+> ⚠️ **OUTPUT REMINDER**: Backend API designs go directly into the output file. DO NOT display API lists or processing logic in conversation.
+
 > **CRITICAL CONSTRAINT**: Backend design in Feature Spec must remain at the BUSINESS API level:
 > - Describe API endpoints as business operations (e.g., "Create Shop" operation with business parameters)
 > - Describe business validation rules in plain language (e.g., "Shop name must be unique within tenant")
@@ -334,6 +371,8 @@ Document: `User Action → Frontend Response → Backend API Call`
 
 ## Step 3: Data Model & Business Rules
 
+> ⚠️ **OUTPUT REMINDER**: Data models and business rules go directly into the output file. DO NOT display tables or validation rules in conversation.
+
 ### 3.1 New Data Structures
 
 | Field | Type | Constraints | Description |
@@ -462,8 +501,8 @@ Document: `User Action → Frontend Response → Backend API Call`
 After user confirms (or if skipped):
 
 ```bash
-node speccrew-workspace/scripts/update-progress.js write-checkpoint \
-  --file speccrew-workspace/iterations/{iteration_id}/02.feature-design/.checkpoints.json \
+node "{workspace_path}/scripts/update-progress.js" write-checkpoint \
+  --file "{workspace_path}/iterations/{iteration_id}/02.feature-design/.checkpoints.json" \
   --stage 02_feature_design \
   --checkpoint feature_design_review \
   --passed true
@@ -623,8 +662,8 @@ Verify all Mermaid diagrams follow compliance rules:
 Set final checkpoint status:
 
 ```bash
-node speccrew-workspace/scripts/update-progress.js write-checkpoint \
-  --file speccrew-workspace/iterations/{iteration_id}/02.feature-design/.checkpoints.json \
+node "{workspace_path}/scripts/update-progress.js" write-checkpoint \
+  --file "{workspace_path}/iterations/{iteration_id}/02.feature-design/.checkpoints.json" \
   --stage 02_feature_design \
   --checkpoint feature_spec_review \
   --passed true

package/.speccrew/skills/speccrew-fd-feature-design/workflow.agentflow.xml

@@ -202,6 +202,14 @@ Does this function breakdown align with your understanding of the requirements?<
        ============================================================ -->
   <sequence id="S1" name="Feature Design" status="pending" desc="Generate Feature Spec from analysis results">
 
+    <!-- Output Efficiency Rule -->
+    <block type="rule" id="R-OUTPUT-EFFICIENCY" level="mandatory" desc="Direct-to-file output only">
+      <field name="text">DO NOT display design content in conversation</field>
+      <field name="text">Think internally, write directly to output file</field>
+      <field name="text">Only output brief status messages and block announcements</field>
+      <field name="text">FORBIDDEN: wireframes, API lists, data tables in chat</field>
+    </block>
+
     <!-- Step 1.0: Prepare analysis results for design -->
     <block type="task" id="B1_0" action="analyze" desc="Prepare analysis data for design phase">
       <field name="input" value="${analyzed_functions}">${prd_content}${feature_id}${feature_name}${feature_type}${frontend_platforms}</field>

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "speccrew",
-  "version": "0.7.31",
+  "version": "0.7.33",
   "description": "Spec-Driven Development toolkit for AI-powered IDEs",
   "author": "charlesmu99",
   "repository": {

package/workspace-template/scripts/update-progress.js

@@ -35,6 +35,7 @@
  *      --output <text>         Task output (used when completed)
  *      --error <text>          Error message (used when failed)
  *      --error-category <cat>  Error category (used when failed)
+ *      --metadata <json>       Metadata JSON string to merge into task
  *
  * 4. update-counts - Force recalculate counts
  *    node update-progress.js update-counts --file <path>
@@ -252,7 +253,8 @@ function parseArgs() {
         confirmedAt: null,
         featuresDir: null,
         platforms: null,
-        force: false
+        force: false,
+        metadata: null
     };
 
     // First argument is the command
@@ -357,6 +359,10 @@ function parseArgs() {
             case '-Matcher-Result':
                 result.matcherResult = args[++i];
                 break;
+            case '--metadata':
+            case '-Metadata':
+                result.metadata = args[++i];
+                break;
         }
     }
 
@@ -619,6 +625,16 @@ function cmdUpdateTask(args) {
             }
         }
 
+        // Handle metadata (merge into task)
+        if (args.metadata) {
+            try {
+                const metadataObj = JSON.parse(args.metadata);
+                task.metadata = { ...task.metadata, ...metadataObj };
+            } catch (e) {
+                outputError(`Failed to parse metadata JSON: ${e.message}`);
+            }
+        }
+
         // Update task
         data.tasks[taskIndex] = task;
         data.updated_at = now;