Project
Workflow Automator
A CLI and API toolkit that connects issue trackers, chat, and CI pipelines into repeatable automation recipes.
- Node.js
- Python
- DevOps
Overview
Workflow Automator reduces manual toil by encoding team workflows as versioned, testable automation scripts. It bridges issue trackers, chat, and CI so the same recipe can run from a pull request, a scheduled job, or a slash command.
Problem
Release checklists, triage handoffs, and status pings are still copy-pasted across tools. When the process lives only in someone’s head — or a wiki page no one updates — quality depends on memory and free time.
I wanted automation that teams can own: recipes in git, reviewable like code, with clear inputs and idempotent steps.
Approach
- Recipes as config — declarative steps with typed inputs and secrets handling
- Adapters for common systems (issues, chat, CI) behind a small interface
- CLI-first for local dry-runs, plus an API for server-side triggers
- Node.js and Python runtimes so teams can extend in the language they already use
Each run produces a structured log so failures are debuggable instead of “it just stopped.”
Highlights
- Dry-run mode that previews side effects before writing anything
- Shared recipe library with version pins per team
- Chat notifications that link back to the run artifact
- Retry policies and circuit breakers for flaky third-party APIs
Outcome
Teams replaced fragile tribal checklists with reviewed automation. Onboarding improved — new hires run the same recipes seniors do — and incident postmortems stopped blaming “someone forgot the manual step.”