TicketSlayer AI

The Problem It Solves Engineering teams lose countless hours manually parsing through vague bug reports in Slack channels ("hey the login button is broken on iphone"), trying to figure out what the problem is, and then manually opening GitHub to create a ticket. TicketSlayer AI eliminates this friction entirely.

How We Built It We built the backend in Python using the slack_bolt framework, leveraging Socket Mode for secure, real-time event listening without needing exposed webhooks. For the AI Brain, we utilized the OpenAI Python SDK routed through the Groq API, utilizing their lightning-fast llama-3.3-70b-versatile model to act as a Senior QA Engineer. Finally, we integrated the GitHub REST API to securely push the formatted JSON payloads directly into our repository as live Issues, creating a seamless Chat-to-Action pipeline.

🛠 Hackathon Requirement Fulfilled: MCP Server Integration To maximize the impact and reach of TicketSlayer AI, we integrated the Model Context Protocol (MCP) by exposing our core Slack-to-GitHub LLM pipeline as a native MCP Server (mcp_server.py). By running this FastMCP server, any external AI Assistant (like Claude Desktop) can connect to TicketSlayer and instantly trigger bug parsing and issue creation via our native create_ticket_from_slack_bug tool over stdio. This modular architecture allows our custom Groq logic to be consumed directly by frontier models!

What We Learned We learned how to orchestrate multi-platform API integrations (Slack + Groq/Llama-3 + GitHub) into a single, cohesive Agentic workflow. We also learned how to effectively prompt LLMs to reliably output strict JSON structures for backend processing, and how to modernize standalone apps by turning them into modular tools via the Model Context Protocol (MCP).

Built With

• ai
• and-then-manually-opening-github-to-create-a-ticket.-ticketslayer-ai-eliminates-this-friction-entirely.-##-how-we-built-it-we-built-the-backend-in-python-using-the-`slack-bolt`-framework
• cohesive-agentic-workflow
• creating-a-seamless-chat-to-action-pipeline.-##-what-we-learned-we-learned-how-to-orchestrate-multi-platform-api-integrations-(slack-+-groq/llama-3-+-github)-into-a-single
• github-api
• groq
• leveraging-**socket-mode**-for-secure
• llama-3
• python
• real-time-event-listening-without-needing-exposed-webhooks.-for-the-ai-brain
• slack-api
• trying-to-figure-out-what-the-problem-is
• utilizing-their-lightning-fast-`llama-3.3-70b-versatile`-model-to-act-as-a-senior-qa-engineer.-finally
• we-integrated-the-github-rest-api-to-securely-push-the-formatted-json-payloads-directly-into-our-repository-as-live-issues
• we-utilized-the-openai-python-sdk-routed-through-the-**groq-api**