This section details on the original issue you should resolve

<issue_title>PRD: Multi-Agent Impact Analysis Framework for Planning Phase</issue_title>
<issue_description>## Product Requirements Document

Background

Currently, agent instructions in downstream projects (e.g., Qwiq) only document build configuration impact analysis. However, comprehensive change planning requires impact analysis across multiple dimensions: code structure, architecture, quality, security, and operations.

Reference: Qwiq PR rjmurillo/ai-agents#136 Discussion

Problem Statement

When planning changes, agents lack a structured framework for:

1. Identifying which specialized agents should review proposed changes
2. Conducting domain-specific impact analysis before implementation
3. Ensuring all affected areas (code, architecture, quality, security, operations) are considered
4. Documenting consultation results in a consistent format

This leads to:

• Implementation without adequate cross-domain review
• Missed downstream impacts (e.g., CI/CD breaks, security gaps, quality regressions)
• Rework and delays when issues are discovered late
• Inconsistent planning quality across projects

Proposed Solution

Add a Multi-Agent Impact Analysis Framework to the ai-agents system that enables the planner agent to orchestrate specialized agent consultations during the planning phase.

Core Capabilities

1. Agent Consultation Protocol

When planner agent creates a plan, it should:

1. Analyze the proposed change scope
2. Identify required specialist agent consultations
3. Invoke each specialist agent with structured impact analysis prompts
4. Aggregate consultation results
5. Document findings in the plan

2. Specialist Agent Roles

3. Consultation Workflow

User Request → planner (scope analysis)
                   ↓
            [Identify required consultations]
                   ↓
         ┌─────────┴──────────┐
         ↓                    ↓
  implementer              architect
  (code impact)     (domain consistency)
         ↓                    ↓
    ┌────┴────┐          ┌───┴───┐
    ↓         ↓          ↓       ↓
   qa      security   devops  [other agents]
         ↓
  [Aggregate findings]
         ↓
  Final Plan Document

4. Impact Analysis Template

Each consultation should produce:

## {Agent} Impact Analysis

**Scope Reviewed**: [Brief description]

### Identified Impacts
- [ ] Impact 1: [Description] - Severity: High/Medium/Low
- [ ] Impact 2: [Description] - Severity: High/Medium/Low

### Affected Areas
- File/Component: [Details]
- File/Component: [Details]

### Recommendations
1. [Recommendation with rationale]
2. [Recommendation with rationale]

### Risks
- Risk 1: [Description + Mitigation]
- Risk 2: [Description + Mitigation]

### Follow-up Actions
- [ ] Action 1: [Assigned agent]
- [ ] Action 2: [Assigned agent]

User Stories

Story 1: Planner Orchestrates Multi-Agent Consultation

As a planner agent
I want to automatically identify and consult specialist agents
So that proposed changes are reviewed across all affected domains

Acceptance Criteria:

• Given a change request scope, identify required specialist consultations
• Invoke each specialist agent with structured prompts
• Aggregate consultation results into final plan
• Mark plan as "consultation-complete" when all specialists have reviewed

Story 2: Implementer Provides Code Impact Analysis

As an implementer agent
I want to analyze code-level impacts of proposed changes
So that architectural patterns, maintainability, and breaking changes are identified early

Acceptance Criteria:

• Identify all files/modules requiring changes
• Assess implementation comp...