About This Session
Modern developers are surrounded by AI tools - but most teams are still using them like smarter chatbots. What if instead of asking AI questions, we built AI skills that automate the most annoying parts of our work? In this light & fun lightning talk, I’ll show how small Claude Code skills can eliminate common engineering pain points by connecting AI to the systems we already use: Slack, GitHub, incident tools, and project trackers. Here are seven skills I built that dramatically improved my workflow and eliminated some of the most repetitive parts of my job: - Slack Thread Archaeologist - extract decisions buried in long Slack threads - PTO Time Machine - instantly catch up on everything that happened while you were away - PR Reviewer Personality Mode - review code like a security engineer, performance expert, or strict maintainer - Postmortem Autopilot - generate incident postmortems from alerts, logs, and timelines - Who Owns This? - identify the right engineer for any piece of code - The 3AM Pager Helper - diagnose incidents faster during outages - What Actually Shipped This Week - generate automatic engineering updates for teams and leadership Each skill is small, composable, and designed to solve a real workflow problem developers face every day. My plan would be to give a quick demo of two of these skills and show the remaining five to the audience. You’ll leave with concrete ideas for building your own AI-powered skills to automate engineering work inside your organization.
Topics
- AI Coding Assistants
- Anthropic
- APIs
- Agentic AI
- Automation
- Product Management
- Productivity