@hustle-together/api-dev-tools 3.6.3 → 3.6.4

package/hooks/enforce-schema.py

@@ -3,7 +3,7 @@
 Hook: PreToolUse for Write/Edit
 Purpose: Block writing implementation if schema not reviewed WITH USER CONFIRMATION
 
-Phase 5 requires:
+Phase 6 requires:
   1. Create Zod schemas based on interview + research
   2. SHOW schema to user in formatted display
   3. USE AskUserQuestion: "Schema matches interview? [Y/n]"
@@ -97,7 +97,7 @@ def main():
 
         print(json.dumps({
             "permissionDecision": "deny",
-            "reason": f"""❌ BLOCKED: Schema creation (Phase 5) not complete.
+            "reason": f"""❌ BLOCKED: Schema creation (Phase 6) not complete.
 
 Status: {status}
 Schema shown: {schema_shown}
@@ -157,7 +157,7 @@ REQUIRED STEPS:
 
 5. If user says "restart":
    • Reset interview phase
-   • Go back to Phase 3
+   • Go back to Phase 4
 
 6. If user says "confirm":
    • Set schema_creation.user_confirmed = true

package/hooks/enforce-scope.py

@@ -3,7 +3,7 @@
 Hook: PreToolUse for Write/Edit
 Purpose: Block writing API code if scope not confirmed BY USER
 
-Phase 1 requires:
+Phase 2 requires:
   1. Present scope understanding to user
   2. USE AskUserQuestion: "Is this correct? [Y/n]"
   3. Record any modifications user requests
@@ -87,7 +87,7 @@ def main():
 
         print(json.dumps({
             "permissionDecision": "deny",
-            "reason": f"""❌ BLOCKED: Scope confirmation (Phase 1) not complete.
+            "reason": f"""❌ BLOCKED: Scope confirmation (Phase 2) not complete.
 
 Status: {status}
 User question asked: {user_question_asked}

package/hooks/enforce-tdd-red.py

@@ -3,7 +3,7 @@
 Hook: PreToolUse for Write/Edit
 Purpose: Block writing implementation if test matrix not approved WITH USER CONFIRMATION
 
-Phase 7 (TDD Red) requires:
+Phase 8 (TDD Red) requires:
   1. Propose test matrix based on interview + schema
   2. SHOW test plan to user (scenarios, edge cases, coverage)
   3. USE AskUserQuestion: "Test plan looks good? [Y/n]"
@@ -144,7 +144,7 @@ Example test structure:
 
         print(json.dumps({
             "permissionDecision": "deny",
-            "reason": f"""❌ BLOCKED: TDD Red phase (Phase 7) not complete.
+            "reason": f"""❌ BLOCKED: TDD Red phase (Phase 8) not complete.
 
 Current status: {tdd_red_status}
 Test count: {test_count}

package/hooks/enforce-verify.py

@@ -3,12 +3,12 @@
 Hook: PreToolUse for Write/Edit
 Purpose: Block refactoring until verification complete WITH USER GAP DECISION
 
-Phase 9 (Verify) requires:
+Phase 10 (Verify) requires:
   1. Re-read original documentation (after tests pass)
   2. Compare implementation to docs - find gaps
   3. SHOW gap analysis to user
   4. USE AskUserQuestion: "Fix gaps? [Y] / Skip? [n]"
-  5. Loop back to Phase 7 if user wants fixes
+  5. Loop back to Phase 8 if user wants fixes
   6. Only proceed to refactor when user decides
 
 Returns:
@@ -94,7 +94,7 @@ def main():
 
         print(json.dumps({
             "permissionDecision": "deny",
-            "reason": f"""❌ BLOCKED: Verification (Phase 9) not complete.
+            "reason": f"""❌ BLOCKED: Verification (Phase 10) not complete.
 
 Status: {status}
 Re-research done: {re_research_done}
@@ -116,7 +116,7 @@ MISSING:
 REQUIRED STEPS:
 
 1. Re-read the ORIGINAL API documentation:
-   • Use Context7 or WebSearch with SAME queries from Phase 2
+   • Use Context7 or WebSearch with SAME queries from Phase 3
    • Compare EVERY documented feature to your implementation
    • Don't rely on memory - actually re-read the docs
 
@@ -146,7 +146,7 @@ REQUIRED STEPS:
    ]
 
 4. If user says "fix":
-   • Loop back to Phase 7 (TDD Red)
+   • Loop back to Phase 8 (TDD Red)
    • Write new tests for missing features
    • Implement and verify again
    • REPEAT until no gaps or user says skip

package/hooks/verify-after-green.py

@@ -1,7 +1,7 @@
 #!/usr/bin/env python3
 """
 Hook: PostToolUse (after test runs)
-Purpose: Trigger Phase 9 (Verify) + Manifest Generation after tests pass
+Purpose: Trigger Phase 10 (Verify) + Manifest Generation after tests pass
 
 This hook detects when tests pass (TDD Green phase complete) and:
   1. Runs the programmatic manifest generation scripts
@@ -243,7 +243,7 @@ def main():
         context_parts.append("---")
         context_parts.append("")
 
-    context_parts.append("## Phase 9: Implementation Verification Required")
+    context_parts.append("## Phase 10: Implementation Verification Required")
     context_parts.append("")
     context_parts.append("Tests are passing. Before proceeding, you MUST verify your implementation:")
     context_parts.append("")

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@hustle-together/api-dev-tools",
-  "version": "3.6.3",
+  "version": "3.6.4",
   "description": "Interview-driven, research-first API development workflow with continuous verification loops for Claude Code",
   "main": "bin/cli.js",
   "bin": {

package/templates/CLAUDE-SECTION.md

@@ -16,19 +16,19 @@ This project uses **@hustle-together/api-dev-tools** for interview-driven, resea
 ### 13-Phase Flow
 
 ```
-Phase 0:  DISAMBIGUATION     - Clarify ambiguous terms before research
-Phase 1:  SCOPE              - Confirm understanding of endpoint
-Phase 2:  INITIAL RESEARCH   - 2-3 targeted searches (Context7, WebSearch)
-Phase 3:  INTERVIEW          - Questions generated FROM discovered params
-Phase 4:  DEEP RESEARCH      - Propose additional searches based on answers
-Phase 5:  SCHEMA             - Create Zod schema from research + interview
-Phase 6:  ENVIRONMENT        - Verify API keys exist
-Phase 7:  TDD RED            - Write failing tests from schema
-Phase 8:  TDD GREEN          - Minimal implementation to pass tests
-Phase 9:  VERIFY             - Re-research docs, compare to implementation
-Phase 10: TDD REFACTOR       - Clean up code while tests pass
-Phase 11: DOCUMENTATION      - Update manifests, cache research
-Phase 12: COMPLETION         - Final verification, commit
+Phase 1:  DISAMBIGUATION     - Clarify ambiguous terms before research
+Phase 2:  SCOPE              - Confirm understanding of endpoint
+Phase 3:  INITIAL RESEARCH   - 2-3 targeted searches (Context7, WebSearch)
+Phase 4:  INTERVIEW          - Questions generated FROM discovered params
+Phase 5:  DEEP RESEARCH      - Propose additional searches based on answers
+Phase 6:  SCHEMA             - Create Zod schema from research + interview
+Phase 7:  ENVIRONMENT        - Verify API keys exist
+Phase 8:  TDD RED            - Write failing tests from schema
+Phase 9:  TDD GREEN          - Minimal implementation to pass tests
+Phase 10: VERIFY             - Re-research docs, compare to implementation
+Phase 11: TDD REFACTOR       - Clean up code while tests pass
+Phase 12: DOCUMENTATION      - Update manifests, cache research
+Phase 13: COMPLETION         - Final verification, commit
 ```
 
 ### Key Principles