@heyai-rules/pilo-masterkit 1.2.2

package/.agent/workflows/multi-plan.md

@@ -0,0 +1,268 @@
+# Plan - Multi-Model Collaborative Planning
+
+Multi-model collaborative planning - Context retrieval + Dual-model analysis → Generate step-by-step implementation plan.
+
+$ARGUMENTS
+
+---
+
+## Core Protocols
+
+- **Language Protocol**: Use **English** when interacting with tools/models, communicate with user in their language
+- **Mandatory Parallel**: Codex/Gemini calls MUST use `run_in_background: true` (including single model calls, to avoid blocking main thread)
+- **Code Sovereignty**: External models have **zero filesystem write access**, all modifications by Claude
+- **Stop-Loss Mechanism**: Do not proceed to next phase until current phase output is validated
+- **Planning Only**: This command allows reading context and writing to `.claude/plan/*` plan files, but **NEVER modify production code**
+
+---
+
+## Multi-Model Call Specification
+
+**Call Syntax** (parallel: use `run_in_background: true`):
+
+```
+Bash({
+  command: "~/.claude/bin/codeagent-wrapper {{LITE_MODE_FLAG}}--backend <codex|gemini> {{GEMINI_MODEL_FLAG}}- \"$PWD\" <<'EOF'
+ROLE_FILE: <role prompt path>
+<TASK>
+Requirement: <enhanced requirement>
+Context: <retrieved project context>
+</TASK>
+OUTPUT: Step-by-step implementation plan with pseudo-code. DO NOT modify any files.
+EOF",
+  run_in_background: true,
+  timeout: 3600000,
+  description: "Brief description"
+})
+```
+
+**Model Parameter Notes**:
+- `{{GEMINI_MODEL_FLAG}}`: When using `--backend gemini`, replace with `--gemini-model gemini-3-pro-preview` (note trailing space); use empty string for codex
+
+**Role Prompts**:
+
+| Phase | Codex | Gemini |
+|-------|-------|--------|
+| Analysis | `~/.claude/.ccg/prompts/codex/analyzer.md` | `~/.claude/.ccg/prompts/gemini/analyzer.md` |
+| Planning | `~/.claude/.ccg/prompts/codex/architect.md` | `~/.claude/.ccg/prompts/gemini/architect.md` |
+
+**Session Reuse**: Each call returns `SESSION_ID: xxx` (typically output by wrapper), **MUST save** for subsequent `/ccg:execute` use.
+
+**Wait for Background Tasks** (max timeout 600000ms = 10 minutes):
+
+```
+TaskOutput({ task_id: "<task_id>", block: true, timeout: 600000 })
+```
+
+**IMPORTANT**:
+- Must specify `timeout: 600000`, otherwise default 30 seconds will cause premature timeout
+- If still incomplete after 10 minutes, continue polling with `TaskOutput`, **NEVER kill the process**
+- If waiting is skipped due to timeout, **MUST call `AskUserQuestion` to ask user whether to continue waiting or kill task**
+
+---
+
+## Execution Workflow
+
+**Planning Task**: $ARGUMENTS
+
+### Phase 1: Full Context Retrieval
+
+`[Mode: Research]`
+
+#### 1.1 Prompt Enhancement (MUST execute first)
+
+**If ace-tool MCP is available**, call `mcp__ace-tool__enhance_prompt` tool:
+
+```
+mcp__ace-tool__enhance_prompt({
+  prompt: "$ARGUMENTS",
+  conversation_history: "<last 5-10 conversation turns>",
+  project_root_path: "$PWD"
+})
+```
+
+Wait for enhanced prompt, **replace original $ARGUMENTS with enhanced result** for all subsequent phases.
+
+**If ace-tool MCP is NOT available**: Skip this step and use the original `$ARGUMENTS` as-is for all subsequent phases.
+
+#### 1.2 Context Retrieval
+
+**If ace-tool MCP is available**, call `mcp__ace-tool__search_context` tool:
+
+```
+mcp__ace-tool__search_context({
+  query: "<semantic query based on enhanced requirement>",
+  project_root_path: "$PWD"
+})
+```
+
+- Build semantic query using natural language (Where/What/How)
+- **NEVER answer based on assumptions**
+
+**If ace-tool MCP is NOT available**, use Claude Code built-in tools as fallback:
+1. **Glob**: Find relevant files by pattern (e.g., `Glob("**/*.ts")`, `Glob("src/**/*.py")`)
+2. **Grep**: Search for key symbols, function names, class definitions (e.g., `Grep("className|functionName")`)
+3. **Read**: Read the discovered files to gather complete context
+4. **Task (Explore agent)**: For deeper exploration, use `Task` with `subagent_type: "Explore"` to search across the codebase
+
+#### 1.3 Completeness Check
+
+- Must obtain **complete definitions and signatures** for relevant classes, functions, variables
+- If context insufficient, trigger **recursive retrieval**
+- Prioritize output: entry file + line number + key symbol name; add minimal code snippets only when necessary to resolve ambiguity
+
+#### 1.4 Requirement Alignment
+
+- If requirements still have ambiguity, **MUST** output guiding questions for user
+- Until requirement boundaries are clear (no omissions, no redundancy)
+
+### Phase 2: Multi-Model Collaborative Analysis
+
+`[Mode: Analysis]`
+
+#### 2.1 Distribute Inputs
+
+**Parallel call** Codex and Gemini (`run_in_background: true`):
+
+Distribute **original requirement** (without preset opinions) to both models:
+
+1. **Codex Backend Analysis**:
+   - ROLE_FILE: `~/.claude/.ccg/prompts/codex/analyzer.md`
+   - Focus: Technical feasibility, architecture impact, performance considerations, potential risks
+   - OUTPUT: Multi-perspective solutions + pros/cons analysis
+
+2. **Gemini Frontend Analysis**:
+   - ROLE_FILE: `~/.claude/.ccg/prompts/gemini/analyzer.md`
+   - Focus: UI/UX impact, user experience, visual design
+   - OUTPUT: Multi-perspective solutions + pros/cons analysis
+
+Wait for both models' complete results with `TaskOutput`. **Save SESSION_ID** (`CODEX_SESSION` and `GEMINI_SESSION`).
+
+#### 2.2 Cross-Validation
+
+Integrate perspectives and iterate for optimization:
+
+1. **Identify consensus** (strong signal)
+2. **Identify divergence** (needs weighing)
+3. **Complementary strengths**: Backend logic follows Codex, Frontend design follows Gemini
+4. **Logical reasoning**: Eliminate logical gaps in solutions
+
+#### 2.3 (Optional but Recommended) Dual-Model Plan Draft
+
+To reduce risk of omissions in Claude's synthesized plan, can parallel have both models output "plan drafts" (still **NOT allowed** to modify files):
+
+1. **Codex Plan Draft** (Backend authority):
+   - ROLE_FILE: `~/.claude/.ccg/prompts/codex/architect.md`
+   - OUTPUT: Step-by-step plan + pseudo-code (focus: data flow/edge cases/error handling/test strategy)
+
+2. **Gemini Plan Draft** (Frontend authority):
+   - ROLE_FILE: `~/.claude/.ccg/prompts/gemini/architect.md`
+   - OUTPUT: Step-by-step plan + pseudo-code (focus: information architecture/interaction/accessibility/visual consistency)
+
+Wait for both models' complete results with `TaskOutput`, record key differences in their suggestions.
+
+#### 2.4 Generate Implementation Plan (Claude Final Version)
+
+Synthesize both analyses, generate **Step-by-step Implementation Plan**:
+
+```markdown
+## Implementation Plan: <Task Name>
+
+### Task Type
+- [ ] Frontend (→ Gemini)
+- [ ] Backend (→ Codex)
+- [ ] Fullstack (→ Parallel)
+
+### Technical Solution
+<Optimal solution synthesized from Codex + Gemini analysis>
+
+### Implementation Steps
+1. <Step 1> - Expected deliverable
+2. <Step 2> - Expected deliverable
+...
+
+### Key Files
+| File | Operation | Description |
+|------|-----------|-------------|
+| path/to/file.ts:L10-L50 | Modify | Description |
+
+### Risks and Mitigation
+| Risk | Mitigation |
+|------|------------|
+
+### SESSION_ID (for /ccg:execute use)
+- CODEX_SESSION: <session_id>
+- GEMINI_SESSION: <session_id>
+```
+
+### Phase 2 End: Plan Delivery (Not Execution)
+
+**`/ccg:plan` responsibilities end here, MUST execute the following actions**:
+
+1. Present complete implementation plan to user (including pseudo-code)
+2. Save plan to `.claude/plan/<feature-name>.md` (extract feature name from requirement, e.g., `user-auth`, `payment-module`)
+3. Output prompt in **bold text** (MUST use actual saved file path):
+
+   ---
+   **Plan generated and saved to `.claude/plan/actual-feature-name.md`**
+
+   **Please review the plan above. You can:**
+   - **Modify plan**: Tell me what needs adjustment, I'll update the plan
+   - **Execute plan**: Copy the following command to a new session
+
+   ```
+   /ccg:execute .claude/plan/actual-feature-name.md
+   ```
+   ---
+
+   **NOTE**: The `actual-feature-name.md` above MUST be replaced with the actual saved filename!
+
+4. **Immediately terminate current response** (Stop here. No more tool calls.)
+
+**ABSOLUTELY FORBIDDEN**:
+- Ask user "Y/N" then auto-execute (execution is `/ccg:execute`'s responsibility)
+- Any write operations to production code
+- Automatically call `/ccg:execute` or any implementation actions
+- Continue triggering model calls when user hasn't explicitly requested modifications
+
+---
+
+## Plan Saving
+
+After planning completes, save plan to:
+
+- **First planning**: `.claude/plan/<feature-name>.md`
+- **Iteration versions**: `.claude/plan/<feature-name>-v2.md`, `.claude/plan/<feature-name>-v3.md`...
+
+Plan file write should complete before presenting plan to user.
+
+---
+
+## Plan Modification Flow
+
+If user requests plan modifications:
+
+1. Adjust plan content based on user feedback
+2. Update `.claude/plan/<feature-name>.md` file
+3. Re-present modified plan
+4. Prompt user to review or execute again
+
+---
+
+## Next Steps
+
+After user approves, **manually** execute:
+
+```bash
+/ccg:execute .claude/plan/<feature-name>.md
+```
+
+---
+
+## Key Rules
+
+1. **Plan only, no implementation** – This command does not execute any code changes
+2. **No Y/N prompts** – Only present plan, let user decide next steps
+3. **Trust Rules** – Backend follows Codex, Frontend follows Gemini
+4. External models have **zero filesystem write access**
+5. **SESSION_ID Handoff** – Plan must include `CODEX_SESSION` / `GEMINI_SESSION` at end (for `/ccg:execute resume <SESSION_ID>` use)

package/.agent/workflows/multi-workflow.md

@@ -0,0 +1,191 @@
+# Workflow - Multi-Model Collaborative Development
+
+Multi-model collaborative development workflow (Research → Ideation → Plan → Execute → Optimize → Review), with intelligent routing: Frontend → Gemini, Backend → Codex.
+
+Structured development workflow with quality gates, MCP services, and multi-model collaboration.
+
+## Usage
+
+```bash
+/workflow <task description>
+```
+
+## Context
+
+- Task to develop: $ARGUMENTS
+- Structured 6-phase workflow with quality gates
+- Multi-model collaboration: Codex (backend) + Gemini (frontend) + Claude (orchestration)
+- MCP service integration (ace-tool, optional) for enhanced capabilities
+
+## Your Role
+
+You are the **Orchestrator**, coordinating a multi-model collaborative system (Research → Ideation → Plan → Execute → Optimize → Review). Communicate concisely and professionally for experienced developers.
+
+**Collaborative Models**:
+- **ace-tool MCP** (optional) – Code retrieval + Prompt enhancement
+- **Codex** – Backend logic, algorithms, debugging (**Backend authority, trustworthy**)
+- **Gemini** – Frontend UI/UX, visual design (**Frontend expert, backend opinions for reference only**)
+- **Claude (self)** – Orchestration, planning, execution, delivery
+
+---
+
+## Multi-Model Call Specification
+
+**Call syntax** (parallel: `run_in_background: true`, sequential: `false`):
+
+```
+# New session call
+Bash({
+  command: "~/.claude/bin/codeagent-wrapper {{LITE_MODE_FLAG}}--backend <codex|gemini> {{GEMINI_MODEL_FLAG}}- \"$PWD\" <<'EOF'
+ROLE_FILE: <role prompt path>
+<TASK>
+Requirement: <enhanced requirement (or $ARGUMENTS if not enhanced)>
+Context: <project context and analysis from previous phases>
+</TASK>
+OUTPUT: Expected output format
+EOF",
+  run_in_background: true,
+  timeout: 3600000,
+  description: "Brief description"
+})
+
+# Resume session call
+Bash({
+  command: "~/.claude/bin/codeagent-wrapper {{LITE_MODE_FLAG}}--backend <codex|gemini> {{GEMINI_MODEL_FLAG}}resume <SESSION_ID> - \"$PWD\" <<'EOF'
+ROLE_FILE: <role prompt path>
+<TASK>
+Requirement: <enhanced requirement (or $ARGUMENTS if not enhanced)>
+Context: <project context and analysis from previous phases>
+</TASK>
+OUTPUT: Expected output format
+EOF",
+  run_in_background: true,
+  timeout: 3600000,
+  description: "Brief description"
+})
+```
+
+**Model Parameter Notes**:
+- `{{GEMINI_MODEL_FLAG}}`: When using `--backend gemini`, replace with `--gemini-model gemini-3-pro-preview` (note trailing space); use empty string for codex
+
+**Role Prompts**:
+
+| Phase | Codex | Gemini |
+|-------|-------|--------|
+| Analysis | `~/.claude/.ccg/prompts/codex/analyzer.md` | `~/.claude/.ccg/prompts/gemini/analyzer.md` |
+| Planning | `~/.claude/.ccg/prompts/codex/architect.md` | `~/.claude/.ccg/prompts/gemini/architect.md` |
+| Review | `~/.claude/.ccg/prompts/codex/reviewer.md` | `~/.claude/.ccg/prompts/gemini/reviewer.md` |
+
+**Session Reuse**: Each call returns `SESSION_ID: xxx`, use `resume xxx` subcommand for subsequent phases (note: `resume`, not `--resume`).
+
+**Parallel Calls**: Use `run_in_background: true` to start, wait for results with `TaskOutput`. **Must wait for all models to return before proceeding to next phase**.
+
+**Wait for Background Tasks** (use max timeout 600000ms = 10 minutes):
+
+```
+TaskOutput({ task_id: "<task_id>", block: true, timeout: 600000 })
+```
+
+**IMPORTANT**:
+- Must specify `timeout: 600000`, otherwise default 30 seconds will cause premature timeout.
+- If still incomplete after 10 minutes, continue polling with `TaskOutput`, **NEVER kill the process**.
+- If waiting is skipped due to timeout, **MUST call `AskUserQuestion` to ask user whether to continue waiting or kill task. Never kill directly.**
+
+---
+
+## Communication Guidelines
+
+1. Start responses with mode label `[Mode: X]`, initial is `[Mode: Research]`.
+2. Follow strict sequence: `Research → Ideation → Plan → Execute → Optimize → Review`.
+3. Request user confirmation after each phase completion.
+4. Force stop when score < 7 or user does not approve.
+5. Use `AskUserQuestion` tool for user interaction when needed (e.g., confirmation/selection/approval).
+
+## When to Use External Orchestration
+
+Use external tmux/worktree orchestration when the work must be split across parallel workers that need isolated git state, independent terminals, or separate build/test execution. Use in-process subagents for lightweight analysis, planning, or review where the main session remains the only writer.
+
+```bash
+node scripts/orchestrate-worktrees.js .claude/plan/workflow-e2e-test.json --execute
+```
+
+---
+
+## Execution Workflow
+
+**Task Description**: $ARGUMENTS
+
+### Phase 1: Research & Analysis
+
+`[Mode: Research]` - Understand requirements and gather context:
+
+1. **Prompt Enhancement** (if ace-tool MCP available): Call `mcp__ace-tool__enhance_prompt`, **replace original $ARGUMENTS with enhanced result for all subsequent Codex/Gemini calls**. If unavailable, use `$ARGUMENTS` as-is.
+2. **Context Retrieval** (if ace-tool MCP available): Call `mcp__ace-tool__search_context`. If unavailable, use built-in tools: `Glob` for file discovery, `Grep` for symbol search, `Read` for context gathering, `Task` (Explore agent) for deeper exploration.
+3. **Requirement Completeness Score** (0-10):
+   - Goal clarity (0-3), Expected outcome (0-3), Scope boundaries (0-2), Constraints (0-2)
+   - ≥7: Continue | <7: Stop, ask clarifying questions
+
+### Phase 2: Solution Ideation
+
+`[Mode: Ideation]` - Multi-model parallel analysis:
+
+**Parallel Calls** (`run_in_background: true`):
+- Codex: Use analyzer prompt, output technical feasibility, solutions, risks
+- Gemini: Use analyzer prompt, output UI feasibility, solutions, UX evaluation
+
+Wait for results with `TaskOutput`. **Save SESSION_ID** (`CODEX_SESSION` and `GEMINI_SESSION`).
+
+**Follow the `IMPORTANT` instructions in `Multi-Model Call Specification` above**
+
+Synthesize both analyses, output solution comparison (at least 2 options), wait for user selection.
+
+### Phase 3: Detailed Planning
+
+`[Mode: Plan]` - Multi-model collaborative planning:
+
+**Parallel Calls** (resume session with `resume <SESSION_ID>`):
+- Codex: Use architect prompt + `resume $CODEX_SESSION`, output backend architecture
+- Gemini: Use architect prompt + `resume $GEMINI_SESSION`, output frontend architecture
+
+Wait for results with `TaskOutput`.
+
+**Follow the `IMPORTANT` instructions in `Multi-Model Call Specification` above**
+
+**Claude Synthesis**: Adopt Codex backend plan + Gemini frontend plan, save to `.claude/plan/task-name.md` after user approval.
+
+### Phase 4: Implementation
+
+`[Mode: Execute]` - Code development:
+
+- Strictly follow approved plan
+- Follow existing project code standards
+- Request feedback at key milestones
+
+### Phase 5: Code Optimization
+
+`[Mode: Optimize]` - Multi-model parallel review:
+
+**Parallel Calls**:
+- Codex: Use reviewer prompt, focus on security, performance, error handling
+- Gemini: Use reviewer prompt, focus on accessibility, design consistency
+
+Wait for results with `TaskOutput`. Integrate review feedback, execute optimization after user confirmation.
+
+**Follow the `IMPORTANT` instructions in `Multi-Model Call Specification` above**
+
+### Phase 6: Quality Review
+
+`[Mode: Review]` - Final evaluation:
+
+- Check completion against plan
+- Run tests to verify functionality
+- Report issues and recommendations
+- Request final user confirmation
+
+---
+
+## Key Rules
+
+1. Phase sequence cannot be skipped (unless user explicitly instructs)
+2. External models have **zero filesystem write access**, all modifications by Claude
+3. **Force stop** when score < 7 or user does not approve

package/.agent/workflows/orchestrate.md

@@ -0,0 +1,231 @@
+---
+description: Sequential and tmux/worktree orchestration guidance for multi-agent workflows.
+---
+
+# Orchestrate Command
+
+Sequential agent workflow for complex tasks.
+
+## Usage
+
+`/orchestrate [workflow-type] [task-description]`
+
+## Workflow Types
+
+### feature
+Full feature implementation workflow:
+```
+planner -> tdd-guide -> code-reviewer -> security-reviewer
+```
+
+### bugfix
+Bug investigation and fix workflow:
+```
+planner -> tdd-guide -> code-reviewer
+```
+
+### refactor
+Safe refactoring workflow:
+```
+architect -> code-reviewer -> tdd-guide
+```
+
+### security
+Security-focused review:
+```
+security-reviewer -> code-reviewer -> architect
+```
+
+## Execution Pattern
+
+For each agent in the workflow:
+
+1. **Invoke agent** with context from previous agent
+2. **Collect output** as structured handoff document
+3. **Pass to next agent** in chain
+4. **Aggregate results** into final report
+
+## Handoff Document Format
+
+Between agents, create handoff document:
+
+```markdown
+## HANDOFF: [previous-agent] -> [next-agent]
+
+### Context
+[Summary of what was done]
+
+### Findings
+[Key discoveries or decisions]
+
+### Files Modified
+[List of files touched]
+
+### Open Questions
+[Unresolved items for next agent]
+
+### Recommendations
+[Suggested next steps]
+```
+
+## Example: Feature Workflow
+
+```
+/orchestrate feature "Add user authentication"
+```
+
+Executes:
+
+1. **Planner Agent**
+   - Analyzes requirements
+   - Creates implementation plan
+   - Identifies dependencies
+   - Output: `HANDOFF: planner -> tdd-guide`
+
+2. **TDD Guide Agent**
+   - Reads planner handoff
+   - Writes tests first
+   - Implements to pass tests
+   - Output: `HANDOFF: tdd-guide -> code-reviewer`
+
+3. **Code Reviewer Agent**
+   - Reviews implementation
+   - Checks for issues
+   - Suggests improvements
+   - Output: `HANDOFF: code-reviewer -> security-reviewer`
+
+4. **Security Reviewer Agent**
+   - Security audit
+   - Vulnerability check
+   - Final approval
+   - Output: Final Report
+
+## Final Report Format
+
+```
+ORCHESTRATION REPORT
+====================
+Workflow: feature
+Task: Add user authentication
+Agents: planner -> tdd-guide -> code-reviewer -> security-reviewer
+
+SUMMARY
+-------
+[One paragraph summary]
+
+AGENT OUTPUTS
+-------------
+Planner: [summary]
+TDD Guide: [summary]
+Code Reviewer: [summary]
+Security Reviewer: [summary]
+
+FILES CHANGED
+-------------
+[List all files modified]
+
+TEST RESULTS
+------------
+[Test pass/fail summary]
+
+SECURITY STATUS
+---------------
+[Security findings]
+
+RECOMMENDATION
+--------------
+[SHIP / NEEDS WORK / BLOCKED]
+```
+
+## Parallel Execution
+
+For independent checks, run agents in parallel:
+
+```markdown
+### Parallel Phase
+Run simultaneously:
+- code-reviewer (quality)
+- security-reviewer (security)
+- architect (design)
+
+### Merge Results
+Combine outputs into single report
+```
+
+For external tmux-pane workers with separate git worktrees, use `node scripts/orchestrate-worktrees.js plan.json --execute`. The built-in orchestration pattern stays in-process; the helper is for long-running or cross-harness sessions.
+
+When workers need to see dirty or untracked local files from the main checkout, add `seedPaths` to the plan file. ECC overlays only those selected paths into each worker worktree after `git worktree add`, which keeps the branch isolated while still exposing in-flight local scripts, plans, or docs.
+
+```json
+{
+  "sessionName": "workflow-e2e",
+  "seedPaths": [
+    "scripts/orchestrate-worktrees.js",
+    "scripts/lib/tmux-worktree-orchestrator.js",
+    ".claude/plan/workflow-e2e-test.json"
+  ],
+  "workers": [
+    { "name": "docs", "task": "Update orchestration docs." }
+  ]
+}
+```
+
+To export a control-plane snapshot for a live tmux/worktree session, run:
+
+```bash
+node scripts/orchestration-status.js .claude/plan/workflow-visual-proof.json
+```
+
+The snapshot includes session activity, tmux pane metadata, worker states, objectives, seeded overlays, and recent handoff summaries in JSON form.
+
+## Operator Command-Center Handoff
+
+When the workflow spans multiple sessions, worktrees, or tmux panes, append a control-plane block to the final handoff:
+
+```markdown
+CONTROL PLANE
+-------------
+Sessions:
+- active session ID or alias
+- branch + worktree path for each active worker
+- tmux pane or detached session name when applicable
+
+Diffs:
+- git status summary
+- git diff --stat for touched files
+- merge/conflict risk notes
+
+Approvals:
+- pending user approvals
+- blocked steps awaiting confirmation
+
+Telemetry:
+- last activity timestamp or idle signal
+- estimated token or cost drift
+- policy events raised by hooks or reviewers
+```
+
+This keeps planner, implementer, reviewer, and loop workers legible from the operator surface.
+
+## Arguments
+
+$ARGUMENTS:
+- `feature <description>` - Full feature workflow
+- `bugfix <description>` - Bug fix workflow
+- `refactor <description>` - Refactoring workflow
+- `security <description>` - Security review workflow
+- `custom <agents> <description>` - Custom agent sequence
+
+## Custom Workflow Example
+
+```
+/orchestrate custom "architect,tdd-guide,code-reviewer" "Redesign caching layer"
+```
+
+## Tips
+
+1. **Start with planner** for complex features
+2. **Always include code-reviewer** before merge
+3. **Use security-reviewer** for auth/payment/PII
+4. **Keep handoffs concise** - focus on what next agent needs
+5. **Run verification** between agents if needed