Agentic tooling · Solo build · 2026
Agentic Workflow Toolkit
A suite of AI-agent skills that automate my team's ops and engineering workflows, from daily standups to LLM-prompt review.
Problem
Recurring team and engineering chores, running standups, tracking sprints across Notion and GitHub, keeping platform docs current, and reviewing LLM prompts before they ship, quietly consumed time and drifted out of date.
Approach
I built a set of Claude Code skills, each a focused agent with the right tools wired in: a standup skill that cross-references GitHub activity with Notion tasks and drafts the daily update; a platform-map skill that scans the repo and regenerates the architecture docs; and a prompt-review skill that audits LLM prompts for efficacy and regressions across the extraction pipeline and the chat agents.
Result
Routine ops and documentation now run as one-command agent invocations instead of manual work, and every prompt change gets a consistent review before it ships.