claude-comms 1.0.41 → 1.0.43

package/.agents/tools/agent-job/agent_job.py

@@ -569,6 +569,7 @@ def spawn_job(harness: str, assignment: str) -> dict:
     status: Optional[JobStatus] = None
     log_file = None
     buffered_lines: list[str] = []  # Lines read after job_id but before fork
+    first_lines: list[str] = []  # Capture first few lines for error reporting
 
     try:
         # Read stdout until we get the job ID
@@ -578,6 +579,10 @@ def spawn_job(harness: str, assignment: str) -> dict:
                     break
                 continue
 
+            # Capture first few lines for debugging if we fail
+            if len(first_lines) < 10:
+                first_lines.append(line.rstrip())
+
             # Try to parse as JSON
             event = try_parse_json(line)
 
@@ -660,7 +665,11 @@ def spawn_job(harness: str, assignment: str) -> dict:
         if log_file:
             log_file.close()
 
-        return {"error": "Failed to extract job ID from harness output"}
+        return {
+            "error": "Failed to extract job ID from harness output",
+            "exit_code": process.poll(),
+            "output": first_lines,
+        }
 
     except Exception as e:
         # Clean up on error
@@ -824,7 +833,7 @@ def main():
             result = spawn_job(args.harness, assignment)
 
             if "error" in result:
-                print(json.dumps(result))
+                print(json.dumps(result, indent=2))
                 sys.exit(1)
 
             # Print user-friendly message with status command

package/.claude/commands/uat-refine.md

@@ -17,22 +17,21 @@ You must manage and maintain Todos dynamically, refine Todos after every decisio
 Populate your initial Todos with your step by step WORKFLOW:
 
 [WORKFLOW]
-PLAN: ULTRATHINK about the UI/UX pages, features, user flows involved and the dependencies between. Break down ASSIGNMENT into dependency sequenced test scenarios.
+PLAN: ULTRATHINK about the UI/UX pages, features, user flows involved and the dependencies between. Break down ASSIGNMENT into dependency sequenced test scenarios batches.
 ENABLE: ensure the dev server is running. you may need to kill an old process and start it again `.agents/repo.md`
 
-For each test scenario in the sequence:
-1. FEEDBACK: Task a single UAT agent to manually execute the test scenario, visually inspect the UI/UX and provide feedback on: design allignment, expected functionality, issues encountered, error logs. This UAT agent is a proxy for the user. Their perspecttive matters!
-2. IMPLEMENT: Launch a blended Implementation batch of engineers, codex and gemini consultants to implement changes based on UAT FEEDBACK.
-3. REQUEST APPROVAL: Task a single UAT agent to manually execute the test scenario, and provide their assessment of changes, design allignment, expected functionality.
+For each test scenario batch sequence:
+1. FEEDBACK: Launch a batch of multiple UAT agents and a support designer. Each UAT agent in the batch should be given a focused test scenario or portion of the app to review. they must visually inspect the UI/UX and provide feedback on: design allignment, expected functionality, issues encountered, error logs. These UAT agents are proxy for the user. Their perspecttive matters!
+2. IMPLEMENT: Launch a blended Implementation batch of engineers + codex consultants + gemini consultants  to implement changes based on UAT FEEDBACK.
+3. REQUEST APPROVAL: Task a batch of one or more (depending on the number of impacted areas/changes) UAT agents. Each UAT agent in this batch should manually execute one of the test scenarios/areas impacted, and provide their assessment of changes, design allignment, expected functionality.
  - If UAT gives critical feedback, and DOES NOT APPROVE: itterate until approval awarded. 
- - When UAT gives signoff approval for deployment, continue to the next test scenario
+ - When all UAT agents in the batch give signoff approval for deployment, continue to the next to the next batch of test scenarios
 [/WORKFLOW]
 
 
 ## 🎯 CRITICAL EXECUTION RULES
 
 ### Dependency Laws
-⚡ **NEVER** batch multiple UAT agents together. Focus on one scenario at a time.
 ⚡ **ALWAYS** Take UAT feedback seriously. They are the user's representitive.
 ⚡ **ALWAYS** Add new todos if Approval is not awarded.
 ⚡ **ALWAYS** blend consultants + engineers for implementation/refinement tasks
@@ -80,9 +79,9 @@ Don't refer to prior batch agents by name. Instead, supply reference artifacts t
 
 
 ### Batch Composition Goals
-- **Minimum**: 5 agents per implementation batch
+- **Minimum**: 5 agents per batch
 - **Target**: 8-10 agents including support roles
-- **Include**: Support Architect/Researcher for guidance
+- **Include**: roughly equal number of gemini, codex, and engineer implementers, and a support architect for guidance/coordination.
 - **Follow with**: Verification batch always
 
 ### Phase Management
@@ -95,7 +94,7 @@ Don't refer to prior batch agents by name. Instead, supply reference artifacts t
 
 ## Consultant Agent
 Consultants come in two variants: Codex and Gemini. leverage them both liberally to diversify.
-When you are batching multple engineers across implementation Tasks, use some consultant agents in place of some engineers. Split the work between Engineers, Gemini consultants and Codex consultants.
+For implementation batches, distribute the tasks across engineers, codex consultants, and gemini consultants. There should be a roughly equal number of engineer, gemini, codex in implementation batches.
 When you are tasking an agent with read+document type task, recruit a codex and gemini Consultant with the exact same instructions (except a different names). Include the consultants in the same batch. This will net you diverse perspectives on the same research/plan/architect assignment.
 
 

package/CLAUDE.md

@@ -221,7 +221,7 @@ When you are tasking an agent with read+document type task, recruit a codex and
 - **designer**: Creates UI/UX specifications, component libraries
 - **engineer**: Implements features end-to-end with code, writes comprehensive test suites, and documentation. Reviews code for quality, security, and standards compliance. Ensures all builds, tests, and checks pass. Visually inspects the running application against UI/UX design/guide.
 - **gemini|codex consultant**: diverse multi-hat. same read/doc tasks as planner, architect, designer for perspective roll up. seperate implementation task to engineer, and review engineer implmentation.
-- **uat**: uses browser automation to manually test user flows. Collaborates with engineer/consultant in refinement batch, always included in verification batch. Max 1 UAT agent per batch. 
+- **uat**: uses browser automation to manually test user flows. Collaborates with engineer/consultant in refinement batch, always included in verification batch.
 
 ## Critical Reminders
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "claude-comms",
-  "version": "1.0.41",
+  "version": "1.0.43",
   "description": "NPX installer for Claude Code hooks and observability setup",
   "main": "src/index.js",
   "bin": {