AI Assisted Insight Report
Availability
✅
🔶
From
October 25
December 25
The AI-powered report functionality generates comprehensive analysis reports for your Salesforce projects using advanced language models. This feature provides deep insights into code quality, architecture, and best practices specific to the Flxbl framework.
Overview
The report generator analyzes your codebase through multiple perspectives:
Package architecture and design patterns
Dependencies and coupling between packages
Code quality and technical debt
Flxbl best practices compliance
Security and compliance considerations
Prerequisites
OpenCode is currently only supported on OSX or Linux runtimes. It's not supported for Windows platforms.
For sfp (community) users: These features are available in alpha. You can use npm install -g @flxbl-io/sfp@ai
to access them.
For complete setup and configuration instructions, see Configuring LLM Providers.
Quick Setup
# Install OpenCode CLI
npm install -g opencode-ai
# Configure Anthropic (recommended)
sfp ai auth --provider anthropic --auth
# Verify authentication
sfp ai auth
# Test provider inference
sfp ai check --provider anthropic
Basic Usage
Package Analysis
# Analyze single package
sfp project report --package nextGen
# Analyze multiple packages
sfp project report --package core --package utils --output core-utils-analysis.md
Domain Analysis ( sfp-pro only)
# Analyze all packages in a domain
sfp project report --domain billing --output billing-analysis.md
Provider-Specific Examples
Anthropic (Recommended)
# Uses defaults (provider: anthropic, model: claude-sonnet-4-20250514)
sfp project report --package nextGen --output nextgen-analysis.md
# Specify different model (if needed)
sfp project report --model claude-sonnet-4-20250514 --package core
GitHub Copilot
# Must specify provider explicitly (uses claude-sonnet-4 by default)
sfp project report --provider github-copilot --package rate-changes
# Uses default model for GitHub Copilot
sfp project report --provider github-copilot --domain service
Amazon Bedrock
# Uses default model: anthropic.claude-sonnet-4-20250514-v1:0
sfp project report --provider amazon-bedrock --package core
# Specify different region (if not in environment)
export AWS_REGION=eu-west-1
sfp project report --provider amazon-bedrock --domain billing
Testing Provider Configuration
Before running reports, you can verify your provider setup using the sfp ai check
command:
# Test all configured providers
sfp ai check
# Test specific provider with default model
sfp ai check --provider anthropic
# Test Amazon Bedrock (uses default: anthropic.claude-sonnet-4-20250514-v1:0)
sfp ai check --provider amazon-bedrock
# Test GitHub Copilot (uses default: claude-sonnet-4)
sfp ai check --provider github-copilot
This command will:
Verify authentication is configured
Test model inference capabilities
Report response time and performance
Help troubleshoot configuration issues
Output Format
Reports are generated in Markdown format with the following structure:
Executive Summary - High-level findings and recommendations
Package/Domain Overview - Architecture and design analysis
Dependencies Analysis - Inter-package relationships
Code Quality Insights - Technical debt and improvement opportunities
Recommendations - Prioritized action items
Troubleshooting
OpenCode CLI Not Found
If you see an error about OpenCode CLI not being installed:
# Install OpenCode CLI
npm install -g opencode-ai
# Verify installation
opencode --version
No Authentication Found
# Check current auth status
sfp ai auth
# If not authenticated, set up credentials
sfp ai auth --provider anthropic --auth
Rate Limiting
If you encounter rate limits:
Reduce
--prompt-count
to lower token usageAnalyze smaller scopes (single package vs domain)
Consider using a different model with higher limits
Cost Considerations
Token Usage: Package analysis typically uses 10-30K tokens, domains 30-80K tokens
Models: Default models are optimized for best value and performance
GitHub Copilot: No additional cost if you have Copilot subscription
Amazon Bedrock: Pay-per-use pricing through AWS, check Bedrock pricing in your region
See Also
Configuring LLM Providers - Complete setup guide for AI providers
AI-Powered PR Linter - Automated PR analysis
Duplicate Check - For identifying duplicate components
Compliance Check - For rule-based code analysis
Last updated