# Validate Configuration Command

You are a configuration validation specialist implementing chain prompting for multi-stage validation.

## Command Usage

```
/project:validate-config [config_path] [profile]
```

**Arguments:**
- `config_path` (optional) - Path to configuration file to validate (default: .claude/config/defaults.json)
- `profile` (optional) - Profile to validate: development|production|research

**Examples:**
```bash
# Validate defaults
/project:validate-config

# Validate development profile
/project:validate-config .claude/config/profiles/development.json development

# Validate custom configuration
/project:validate-config examples/custom_config.json

# Validate merged configuration for production profile
/project:validate-config production
```

## Chain Prompting Validation Stages

```
STAGE 1: Load Configuration & Schema
  ↓ <loaded_config>
STAGE 2: Schema Validation
  ↓ <schema_validation>
STAGE 3: Semantic Validation
  ↓ <semantic_validation>
STAGE 4: Cross-Field Validation
  ↓ <cross_field_validation>
STAGE 5: Generate Validation Report
  ↓ <validation_report>
```

## STAGE 1: Load Configuration & Schema

**Single-Task Goal:** Load configuration file and validation schema

### Actions
1. Determine which configuration to validate:
   - If no args: validate defaults.json
   - If profile name only: load defaults.json + profile.json and merge
   - If config_path: load specified file
2. Read configuration file using Read tool
3. Read schema.json using Read tool
4. Parse JSON content

### Output State
```xml
<loaded_config>
  <config_path>{path}</config_path>
  <config_content>{...}</config_content>
  <schema_content>{...}</schema_content>
  <load_status>success|failed</load_status>
  <load_errors>...</load_errors>
</loaded_config>
```

## STAGE 2: Schema Validation

**Single-Task Goal:** Validate configuration against JSON schema

### Validation Checks
1. **Required Fields**
   - Check all required properties present
   - Check nested required fields

2. **Type Checking**
   - Validate all types match schema (string, integer, boolean, etc.)
   - Check array item types
   - Check object property types

3. **Value Constraints**
   - Check minimum/maximum for integers
   - Check enum values
   - Check string patterns (regex)

4. **Structure Validation**
   - Verify object structure matches schema
   - Check for unexpected properties
   - Validate nested objects

### Output State
```xml
<schema_validation>
  <status>passed|failed</status>
  <errors>
    <error>
      <path>orchestration.max_parallel_agents</path>
      <expected>integer between 1 and 10</expected>
      <actual>15</actual>
      <message>Value exceeds maximum allowed</message>
    </error>
    ...
  </errors>
  <warnings>
    <warning>
      <path>generation.custom_field</path>
      <message>Property not defined in schema</message>
    </warning>
  </warnings>
</schema_validation>
```

## STAGE 3: Semantic Validation

**Single-Task Goal:** Validate logical consistency and semantic correctness

### Validation Checks
1. **Logical Consistency**
   - `batch_size` should be <= `max_iterations`
   - `infinite_mode_wave_size` should be reasonable (1-20)
   - `context_budget_per_agent` * `max_parallel_agents` should be < total budget
   - `agent_timeout_ms` should be > `web_fetch_timeout_ms`

2. **Path Validation**
   - `output_directory` should be valid directory name
   - No absolute paths where relative expected
   - No parent directory references (../) in outputs

3. **Value Reasonableness**
   - `max_parallel_agents` not too high (warn if > 10)
   - `batch_size` not too large (warn if > 50)
   - `max_iterations` reasonable for mode
   - Timeout values sensible

4. **Feature Dependencies**
   - If `web_enhancement.enabled` is true, check web-related config
   - If `enable_review_stage` is true, check quality config
   - If `enable_url_strategy` is true, validate URL-related settings

### Output State
```xml
<semantic_validation>
  <status>passed|failed|warning</status>
  <errors>
    <error>
      <check>batch_size_vs_max_iterations</check>
      <message>batch_size (100) exceeds max_iterations (50)</message>
      <severity>error</severity>
    </error>
  </errors>
  <warnings>
    <warning>
      <check>parallel_agents_count</check>
      <message>max_parallel_agents (15) is unusually high, may cause performance issues</message>
      <severity>warning</severity>
    </warning>
  </warnings>
</semantic_validation>
```

## STAGE 4: Cross-Field Validation

**Single-Task Goal:** Validate relationships between configuration sections

### Validation Checks
1. **Orchestration vs Quality**
   - If `enable_review_stage` is true, ensure sufficient timeout
   - Review stage adds 30-50% to execution time

2. **Orchestration vs Limits**
   - `max_parallel_agents` * `infinite_mode_wave_size` should respect `max_iterations`
   - Total context budget calculation: verify feasibility

3. **Web Enhancement vs Orchestration**
   - `initial_priming_urls` should be reasonable for `batch_size`
   - `urls_per_iteration` should be feasible within `agent_timeout_ms`
   - Web cache settings compatible with output settings

4. **Logging vs Performance**
   - Verbose logging with large batch sizes may impact performance
   - Debug logging with many agents may produce excessive output

5. **Chain Prompting vs Quality**
   - If many stages configured, ensure timeout sufficient
   - Self-correction enabled requires retry budget

### Output State
```xml
<cross_field_validation>
  <status>passed|failed|warning</status>
  <checks_performed>
    <check name="orchestration_quality">passed</check>
    <check name="orchestration_limits">warning</check>
    <check name="web_orchestration">passed</check>
    <check name="logging_performance">warning</check>
    <check name="chain_quality">passed</check>
  </checks_performed>
  <issues>
    <issue>
      <check>orchestration_limits</check>
      <fields>
        <field>orchestration.max_parallel_agents</field>
        <field>orchestration.infinite_mode_wave_size</field>
        <field>limits.max_iterations</field>
      </fields>
      <message>Configuration may exceed max_iterations in first wave</message>
      <severity>warning</severity>
      <recommendation>Reduce wave_size or increase max_iterations</recommendation>
    </issue>
  </issues>
</cross_field_validation>
```

## STAGE 5: Generate Validation Report

**Single-Task Goal:** Create comprehensive validation report

### Report Contents
1. **Summary**
   - Overall status: VALID | INVALID | WARNINGS
   - Configuration file validated
   - Profile (if applicable)
   - Validation timestamp

2. **Schema Validation Results**
   - Total checks performed
   - Errors found
   - Warnings found

3. **Semantic Validation Results**
   - Logical consistency checks
   - Value reasonableness
   - Feature dependencies

4. **Cross-Field Validation Results**
   - Relationship checks
   - Compatibility issues

5. **Detailed Error List**
   - All errors with paths and messages
   - Severity levels
   - Recommendations for fixes

6. **Warnings List**
   - All warnings
   - Impact assessment
   - Suggestions for optimization

7. **Configuration Summary**
   - Key settings extracted
   - Profile characteristics
   - Estimated resource usage

### Output Format
```
========================================
CONFIGURATION VALIDATION REPORT
========================================

Configuration: {path}
Profile: {profile}
Validated: {timestamp}
Schema Version: {version}

OVERALL STATUS: {VALID|INVALID|WARNINGS}

----------------------------------------
SCHEMA VALIDATION
----------------------------------------
Status: {passed|failed}
Errors: {n}
Warnings: {n}

[If errors exist:]
ERRORS:
  1. orchestration.max_parallel_agents
     Expected: integer between 1 and 10
     Actual: 15
     Message: Value exceeds maximum allowed

  2. generation.naming_pattern
     Expected: string
     Actual: null
     Message: Required field is missing

[If warnings exist:]
WARNINGS:
  1. generation.custom_field
     Message: Property not defined in schema
     Impact: Will be ignored during execution

----------------------------------------
SEMANTIC VALIDATION
----------------------------------------
Status: {passed|failed|warning}

LOGICAL CONSISTENCY: {passed|failed}
VALUE REASONABLENESS: {passed|warning}
FEATURE DEPENDENCIES: {passed}

[If issues exist:]
ISSUES:
  1. [WARNING] batch_size unusually large
     Value: 100
     Recommendation: Consider reducing to 20-50 for better control

  2. [ERROR] timeout too short for web fetches
     agent_timeout_ms: 10000
     web_fetch_timeout_ms: 30000
     Fix: Increase agent_timeout_ms to at least 60000

----------------------------------------
CROSS-FIELD VALIDATION
----------------------------------------
Status: {passed|warning}

Checks Performed: 5
  ✓ Orchestration vs Quality
  ⚠ Orchestration vs Limits
  ✓ Web Enhancement vs Orchestration
  ⚠ Logging vs Performance
  ✓ Chain Prompting vs Quality

ISSUES:
  1. [WARNING] Configuration may exceed max_iterations
     max_parallel_agents (5) * infinite_mode_wave_size (20) = 100
     max_iterations: 50
     Recommendation: Reduce wave_size to 10 or increase max_iterations

  2. [WARNING] Verbose logging with large batch size
     Impact: May produce excessive output and slow execution
     Recommendation: Use 'info' level instead of 'debug' for batch_size > 10

----------------------------------------
CONFIGURATION SUMMARY
----------------------------------------
Profile: {profile}
Parallel Agents: {n}
Batch Size: {n}
Web Enhancement: {enabled|disabled}
Chain Prompting Stages: {n}
Estimated Context Budget: {tokens} tokens
Estimated Execution Time: {minutes} minutes

----------------------------------------
RECOMMENDATIONS
----------------------------------------
{List of recommended changes based on validation}

1. Reduce max_parallel_agents from 15 to 10
2. Increase agent_timeout_ms to 60000
3. Consider using 'info' logging level instead of 'debug'

========================================
{VALIDATION PASSED|VALIDATION FAILED|VALIDATION PASSED WITH WARNINGS}
========================================
```

## Self-Correction Mode

If validation fails and you want to suggest fixes:

```bash
/project:validate-config <config_path> --suggest-fixes
```

This will:
1. Run all validation stages
2. Identify all errors
3. Generate corrected configuration
4. Display diff showing proposed changes
5. Optionally write corrected config to new file

## Output Examples

### Example 1: Valid Configuration
```
========================================
CONFIGURATION VALIDATION REPORT
========================================

Configuration: .claude/config/profiles/development.json
Profile: development
Validated: 2025-10-10T14:30:00Z
Schema Version: 1.0.0

OVERALL STATUS: VALID

Schema Validation: PASSED (0 errors, 0 warnings)
Semantic Validation: PASSED
Cross-Field Validation: PASSED

Configuration is valid and ready to use.
========================================
```

### Example 2: Configuration with Warnings
```
========================================
CONFIGURATION VALIDATION REPORT
========================================

Configuration: examples/custom_config.json
Profile: custom
Validated: 2025-10-10T14:35:00Z
Schema Version: 1.0.0

OVERALL STATUS: VALID WITH WARNINGS

Schema Validation: PASSED (0 errors, 1 warning)
Semantic Validation: WARNING
Cross-Field Validation: WARNING

WARNINGS:
  1. max_parallel_agents set to 8 (high)
     Impact: May cause performance issues
  2. Debug logging with batch_size 20
     Impact: Excessive output

RECOMMENDATIONS:
  1. Consider reducing max_parallel_agents to 5-6
  2. Use 'info' logging level for production

Configuration is usable but optimizations recommended.
========================================
```

### Example 3: Invalid Configuration
```
========================================
CONFIGURATION VALIDATION REPORT
========================================

Configuration: broken_config.json
Validated: 2025-10-10T14:40:00Z

OVERALL STATUS: INVALID

ERRORS:
  1. orchestration.max_parallel_agents: 15 (exceeds maximum of 10)
  2. generation.naming_pattern: missing required field
  3. agent_timeout_ms (10000) < web_fetch_timeout_ms (30000)

Configuration cannot be used until errors are fixed.

Run with --suggest-fixes to see recommended corrections.
========================================
```

## Exit Codes

When used programmatically:
- Exit 0: Configuration valid
- Exit 1: Configuration invalid (errors)
- Exit 2: Configuration valid with warnings