speccrew 0.7.51 → 0.7.52

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -93,13 +93,23 @@ This skill MUST execute tasks continuously without unnecessary interruptions.
93
93
  3. Unrecoverable errors that prevent further progress
94
94
  4. Security-sensitive operations (e.g., deleting existing files)
95
95
 
96
- ### Phase 3 FORBIDDEN Actions
97
-
98
- - DO NOT inline-call speccrew-sd-framework-evaluate skill
99
- - DO NOT read feature spec files yourself for framework evaluation
100
- - DO NOT create temporary scripts for batch file analysis
101
- - DO NOT generate framework-evaluation.md as fallback if worker fails
102
- - Worker dispatch failure = ABORT (report error, do NOT retry with inline execution)
96
+ ### Phase 3 Execution Method — Worker Agent Dispatch
97
+
98
+ **HOW TO DISPATCH**: When executing Phase 3 (Framework Evaluation), you MUST:
99
+ 1. Start a NEW agent session by invoking `/speccrew-task-worker`
100
+ 2. Pass the following context to the worker:
101
+ - skill_path: ${workspace_path}/.speccrew/skills/speccrew-sd-framework-evaluate/SKILL.md
102
+ - workspace_path, iteration_path, feature_spec_paths, api_contract_paths, techs_knowledge_paths, output_path
103
+ 3. Wait for the worker agent to complete
104
+ 4. Verify that framework-evaluation.md has been generated at the output path
105
+ 5. Then proceed to Phase 3 HARD STOP for user confirmation
106
+
107
+ **FORBIDDEN**:
108
+ - DO NOT use the `Skill` tool to invoke speccrew-sd-framework-evaluate yourself
109
+ - DO NOT read feature spec files or tech knowledge files yourself for framework evaluation
110
+ - DO NOT create any scripts for batch file analysis
111
+ - DO NOT generate framework-evaluation.md yourself under any circumstances
112
+ - If worker agent fails, ABORT — do NOT fallback to inline execution
103
113
 
104
114
  ### HARD STOP Checkpoints
105
115
 
@@ -178,7 +178,7 @@
178
178
 
179
179
  <!-- Step 3.1: Dispatch Worker Agent for Framework Evaluation -->
180
180
  <block type="task" id="P3-B1" action="dispatch-to-worker" status="pending"
181
- desc="Dispatch single worker agent for framework evaluation">
181
+ desc="Dispatch single worker agent for framework evaluation. Use /speccrew-task-worker agent, NOT Skill tool">
182
182
  <field name="agent">speccrew-task-worker</field>
183
183
  <field name="skill">speccrew-sd-framework-evaluate</field>
184
184
  <field name="max_concurrent">1</field>
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "speccrew",
3
- "version": "0.7.51",
3
+ "version": "0.7.52",
4
4
  "description": "Spec-Driven Development toolkit for AI-powered IDEs",
5
5
  "author": "charlesmu99",
6
6
  "repository": {