Feature Requests

Feature Request: Improve Agent & Tool Call Organization with Folders/Tags
As our team's usage of Vapi grows, the flat list of Agents and Tool Calls in the dashboard is becoming increasingly difficult to manage. With multiple developers working on different features and environments (development, staging, production), the list is becoming cluttered and hard to navigate. This lack of organization leads to several challenges: Difficulty in distinguishing environments: It's hard to quickly tell which agent is for production versus a development or testing version. Increased risk of human error: A developer could accidentally edit a live production agent when they intended to modify a development version, potentially causing service disruptions. Inefficient workflow: Finding a specific agent or tool call requires manual searching or relying on strict naming conventions, which is cumbersome and doesn't scale well. Onboarding complexity: New team members have a harder time understanding the structure of our voice agents without a clear organizational hierarchy. I propose introducing a system to group and organize Agents and Tool Calls within the Vapi dashboard. This would provide a much-needed structure for teams and larger projects. Here are a few potential implementations, any of which would be a significant improvement: Folders/Directories: Allow users to create folders to group related agents and tools. For example, we could create folders like Production, Staging, Development, or project-based folders like Customer Support Bot and Sales Outreach Bot. A collapsible folder structure in the sidebar would be an intuitive UI for this. Tagging/Labeling System: Allow users to apply one or more text-based tags to each Agent and Tool Call (e.g., prod, dev, v2, legacy, billing-tool). The dashboard interface would then need a way to filter the list based on these tags. This offers more flexibility than folders as an item can have multiple tags.
0
Advanced Debugging Suite for Vapi Workflows
Vapi.ai workflows function as a visual IDE for voice AI development, but currently lack the comprehensive debugging tools that developers expect from modern development environments. Without proper debugging capabilities, developers struggle to understand workflow execution, troubleshoot issues, and optimize performance. Requested Features: Real-Time Debugging Dashboard Live variable inspector showing current values of all workflow variables during execution Real-time display of LLM extraction results, including confidence scores and alternative interpretations Step-by-step execution flow with visual indicators of current processing stage Call state visualization showing active functions, pending operations, and completion status Breakpoint System Ability to set conditional breakpoints at any workflow node Pause execution when specific variables reach certain values or conditions Resume, step-over, and step-into controls for granular execution control Breakpoint management panel for organizing and toggling multiple breakpoints Execution History & Logging Complete execution trace with timestamps for each workflow step Detailed logs of all LLM interactions, including prompts sent and responses received Variable state snapshots at each execution point Error stack traces with context about where failures occurred Interactive Testing Tools Ability to inject test inputs at any point in the workflow Mock different user responses to test conversation branches Replay previous conversations with modified parameters A/B testing framework for comparing different workflow versions Performance Monitoring Execution time analysis for each workflow component Token usage tracking and cost analysis per workflow run Latency monitoring for LLM calls and API integrations Memory usage and resource consumption metrics Value Proposition: Just as JetBrains IDEs revolutionized code development with intelligent debugging tools, Vapi.ai workflows need equivalent capabilities for voice AI development. This would dramatically reduce development time, improve workflow reliability, and enable developers to build more sophisticated voice applications with confidence. This would position Vapi.ai as the premier platform for professional voice AI development by providing the debugging infrastructure that complex workflows demand.
0
Load More