FEATURE REQUEST: Technical Support Handler with HIL capability #118
Description
Is your feature request related to a problem?
- Yes, it is related to a problem
Describe the feature you'd like
🌟 Feature Description
Implement an Interactive Technical Support Workflow using LangGraph's Human-in-the-Loop (HIL) capabilities. This feature will transform the existing troubleshooting process into a collaborative, real-time debugging session between the Devr.AI agent and the user who reported the issue.
🔍 Problem Statement
Currently, the agent's technical support workflow operates in a linear, non-interactive manner. It attempts to solve a problem by querying its knowledge sources (FalkorDB, Web Search) and providing a final answer. This approach has a significant drawback: if the agent's initial assumptions are wrong, it can waste resources pursuing an incorrect path and ultimately provide an irrelevant or unhelpful solution. It lacks a mechanism to validate its understanding and proposed actions with the person who has the most context—the user.
🎯 Expected Outcome
The successful implementation of this feature will result in a conversational and iterative troubleshooting experience.
- Collaborative Problem-Solving: The agent will work with the user, not just for them. It will pause at critical steps to ask for clarification, confirmation, and guidance.
- Increased Accuracy: By validating its approach with the user, the agent's actions will be far more targeted and relevant, leading to a higher probability of resolving the issue correctly.
- Better User Experience: Users will feel more engaged and in control of the support process, turning potential frustration into a productive partnership.
- Efficient Resource Use: The agent will avoid pursuing incorrect solution paths by getting immediate feedback, saving computational resources and time.
📷 Screenshots and Design Ideas
The core design is a conversational UI where the agent uses LangGraph's interrupt feature to pause and wait for the user's input.
Conceptual Flow:
-
User asks a question: "I'm getting a
ConnectionRefusedErrorin my dev environment." -
Agent Clarifies (HIL Interrupt 1): Instead of immediately searching, the agent asks for context.
Agent: "I can help with that. To start, are you working in a specific repository, like
AOSSIE-Project-X?"
[Workflow Paused: Awaiting User Input] -
User Responds: "Yes, in Project-X."
-
Agent Proposes Action (HIL Interrupt 2): The agent decides to use FalkorDB and seeks approval.
Agent: "Great. I plan to query the
database_connector.pyfile within that repository to check its configuration. Does that sound like the right first step?"
[Workflow Paused: Awaiting User Input] -
User Approves: "Yes, go ahead."
-
Agent Presents Options (HIL Interrupt 3): After analysis, the agent finds multiple possibilities and asks the user to help narrow them down.
Agent: "My analysis suggests the error could be due to one of two things: an incorrect database host in your
.envfile or the Docker container for the DB not being active. Which one should we investigate first?"
[Workflow Paused: Awaiting User Input]
This interactive back-and-forth continues until the user confirms the issue is resolved.
📋 Additional Context
This feature is a direct evolution of our goal to create a truly intelligent and helpful AI-powered DevRel assistant. It moves the agent from a simple "answer machine" to an active, collaborative partner in the developer workflow. This aligns with the core vision of Devr.AI to foster better engagement and provide meaningful support to the open-source community.
Record
- I agree to follow this project's Code of Conduct
- I want to work on implementing this feature