feat(copilot-v1): Copilot v1

[Sg312]

Description

Adds v1 of copilot chat to workflow. Adds internal tools for the copilot agent to access and call. DOES NOT EXPOSE COPILOT IN THIS PR

Fixes # (issue)

Type of change

Please delete options that are not relevant.

• New feature (non-breaking change which adds functionality)
• Documentation update

How Has This Been Tested?

Manual testing of copilot

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have added tests that prove my fix is effective or that my feature works
• All tests pass locally and in CI (bun run test)
• My changes generate no new warnings
• Any dependent changes have been merged and published in downstream modules
• I have updated version numbers as needed (if needed)
• I confirm that I have read and agree to the terms outlined in the Contributor License Agreement (CLA)

Security Considerations:

• My changes do not introduce any new security vulnerabilities
• I have considered the security implications of my changes

Additional Information:

Any additional information, configuration or data that might be necessary to reproduce the issue or use the feature.

[greptile-apps]

Greptile Summary

This PR introduces Copilot v1, a comprehensive AI-powered workflow assistant for Sim Studio. The implementation transforms the platform from a manual workflow builder into an intelligent system that can understand, analyze, and modify workflows through natural language interactions.

Core Architecture

The Copilot system operates in two distinct modes:

• Ask Mode: Provides educational guidance and workflow analysis without editing capabilities
• Agent Mode: Enables direct workflow editing with full AI assistance for building and modifying workflows

Key Components Added

Database Infrastructure: New tables for copilot_chats (conversation storage with JSONB messages) and copilot_checkpoints (workflow state versioning), enabling persistent conversations and rollback capabilities.

YAML Workflow System: Comprehensive YAML documentation and parsing utilities covering all block types (starter, agent, function, api, condition, router, evaluator, response, loop, parallel, webhook, workflow). This includes block reference syntax, connection configurations, and validation rules.

Tool Ecosystem: Four critical internal tools power the AI assistant:

• get_blocks_and_tools: Discovery of available blocks and their capabilities
• get_blocks_metadata: Detailed schema information for specific blocks
• get_yaml_structure: Complete YAML syntax guide and best practices
• edit_workflow: Direct workflow modification through YAML processing

Real-time Features: Enhanced socket server with copilot-workflow-edit and workflow-updated events for live collaboration when AI modifies workflows.

Professional UI: Complete redesign of the chat interface with tool call visualization, checkpoint management panels, mode switching, and streaming message support with inline tool execution feedback.

Auto-layout System: Intelligent workflow visualization with multiple algorithms (hierarchical, smart, force-directed, layered) that automatically arranges blocks based on complexity analysis.

Integration Points

The Copilot integrates deeply with existing workflow infrastructure through consolidated API endpoints (/api/workflows/{id}/yaml, /api/workflows/{id}/state) and extends all LLM providers (Anthropic, OpenAI, Azure, Groq, Cerebras, Deepseek, Ollama, xAI) to pass chatId context for checkpoint operations.

Provider Enhancements

Significant streaming improvements to the Anthropic provider add three execution modes: early streaming (no tools), incremental streaming with real-time tool call feedback (for copilot), and silent tool execution with final response streaming (for chat applications).

Confidence Score: 3/5

• Moderate risk due to several implementation concerns and missing validations that could cause runtime issues
• Major architectural changes with extensive new functionality increase complexity and potential failure points
• Multiple files need attention including API endpoint mismatches, type safety issues, and validation schema problems

_{97 files reviewed, 58 comments}

_{Edit Code Review Bot Settings | Greptile}