docs: add github templates for PR

.github/config_template.md

@@ -0,0 +1,38 @@
+# Configuration Update PR Template
+
+## Configuration Change
+
+### What configuration is being updated?
+<!-- Specify: tasks, models, guardrails, policies, or metrics -->
+
+### Configuration Type
+- [ ] **Tasks** - Task definitions and metadata
+- [ ] **Models** - Model information and capabilities
+- [ ] **Guardrails** - Operational constraints and policies
+- [ ] **Policies** - Policy definitions with embedded thresholds
+- [ ] **Metrics** - Metric definitions and types
+
+---
+
+## Change Details
+
+### What was changed?
+<!-- Brief description of the configuration update -->
+
+### Why was this change needed?
+<!-- One sentence explaining the motivation -->
+
+---
+
+## Validation
+
+### Configuration Validation
+- [ ] All modified files pass `evalguard config validate`
+- [ ] Schema compliance verified against `schemas/v1/`
+- [ ] Configuration follows project structure guidelines
+
+---
+
+## Additional Notes
+
+<!-- Any other relevant information -->

.github/pull_request_template.md

@@ -0,0 +1,26 @@
+# Pull Request Template
+
+## Type of Change
+
+Please select the type of change this PR represents and use the appropriate template:
+
+- [ ] **Report Contribution** - Use [Report Contribution PR Template](.github/report_template.md)
+- [ ] **Configuration Update** - Use [Configuration Template](.github/config_template.md)
+- [ ] **Specification change** - Use [Specification PR Template](.github/spec_template.md)
+- [ ] **Other** - Please describe below
+
+---
+
+## Quick Description
+
+<!-- Provide a brief description of your changes -->
+
+---
+
+## Type-Specific Template
+
+Please copy the content from the appropriate template above and paste it below, then fill it out completely:
+
+```markdown
+<!-- Paste the appropriate template content here -->
+```

.github/report_template.md

@@ -0,0 +1,26 @@
+# Report Contribution PR Template
+
+## Report Information
+
+### Model Details
+- **Model Organization**: <!-- e.g., meta-llama, RedHatAI, microsoft -->
+- **Model Name**: <!-- e.g., Llama-3.1-8B-Instruct, phi-2 -->
+- **Model Version**: <!-- if applicable -->
+- **Report Format**: <!-- e.g., lm-eval, custom format -->
+
+## Checklist
+
+- [ ] I have read and followed the [Contributing Guidelines](README.md#contributing)
+- [ ] Report follows the required directory structure
+- [ ] Report format is supported or discussed in an issue
+- [ ] I have used `evalguard lm-eval gen` to generate missing configuration files
+- [ ] All generated configuration files have been reviewed and enhanced
+- [ ] Configuration files pass validation: `evalguard config validate`
+- [ ] No placeholder text remains in configurations
+- [ ] I have enhanced configurations with meaningful descriptions
+
+---
+
+## Additional Notes
+
+<!-- Any additional information about the evaluation or model -->

.github/spec_template.md

@@ -0,0 +1,24 @@
+# Specification Change PR Template
+
+## Change Summary
+
+### What specification change is being proposed?
+<!-- Brief description of the change to schemas, validation rules, or API spec -->
+
+### Why is this change needed?
+<!-- One sentence explaining the motivation -->
+
+---
+
+## Technical Details
+
+### Change Type
+- [ ] **API Schema** - REST API schema modification
+- [ ] **Config Schema** - Configuration schema
+
+---
+
+## Additional Notes
+
+<!-- Any other relevant information -->
+

README.md

@@ -102,17 +102,18 @@ EvalGuard thrives on community contributions! We welcome contributions of evalua
 #### Report Format
 Reports must follow the directory structure:
 ```
-reports/<model_org>/<model_name>/lm-eval/report.json
+reports/<model_org>/<model_name>/<report_format>/report.json
 ```
 
 **Example**: `reports/meta-llama/Llama-3.1-8B-Instruct/lm-eval/report.json`
 
 #### Supported Formats
-- **lm-evaluation-harness**: Currently supported with full CLI integration
+- **lm-evaluation-harness** (`lm-eval`): Currently supported with full CLI integration
 - **Other formats**: If you have evaluation reports in other formats, please create an issue to discuss integration
 
 #### Adding Missing Data
-When adding a new report, you can use the CLI to automatically generate missing tasks, models, and metrics:
+When adding a new report, you can use the CLI to automatically generate missing tasks, models, and metrics but bear
+in mind that some fields will not be filled like friendly name or description so take some time to review what is missing:
 
 ```bash
 # Generate missing configuration from a single report
@@ -124,6 +125,17 @@ evalguard lm-eval gen -d reports/
 
 This will create the necessary configuration files in the appropriate `config/` subdirectories.
 
+#### Ready to Contribute?
+**🚀 Create a Pull Request** using our [Report Contribution Template](.github/report_template.md) to submit your evaluation reports!
+
+The template will guide you through:
+- Proper report structure and format
+- Configuration generation and validation
+- Quality assurance steps
+- Required checklist items
+
+Simply copy the template content when creating your PR to ensure all requirements are met.
+
 ### Contributing Configuration Files
 
 #### Adding New Configurations
@@ -151,13 +163,16 @@ evalguard config validate -t guardrails
 #### Schema Compliance
 All configuration files must follow the schemas defined in `schemas/v1/`. See [SPECIFICATION.md](SPECIFICATION.md) for detailed schema definitions and examples.
 
-### Contribution Guidelines
-
-1. **Evaluation Reports**: Submit reports following the directory structure `reports/<model_org>/<model_name>/lm-eval/report.json`. Currently, only [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) reports are supported. For other evaluation formats, please create an issue to discuss integration.
+#### Ready to Contribute Configurations?
+**⚙️ Create a Pull Request** using our [Configuration Update Template](.github/config_template.md) to submit your configuration changes!
 
-2. **Configuration Generation**: When contributing new reports, utilize the CLI tool to automatically generate missing tasks, models, and metrics: `evalguard lm-eval gen`
+The template will guide you through:
+- Proper configuration file structure
+- Validation and schema compliance
+- Quality assurance for configuration updates
+- Required checklist items
 
-3. **Configuration Validation**: Ensure all configuration files adhere to the API specification and validate them using: `evalguard config validate`
+Simply copy the template content when creating your PR to ensure all requirements are met.
 
 ---