AI Playground

Slack Feedback Capture Agent

A lightweight service that captures product feedback from Slack, structures it, and makes it ready for review and downstream insight workflows.

View on GitHub

Why it is useful

—Turns ad-hoc product feedback in Slack into a reliable, reviewable pipeline.
—Reduces manual copy/paste from threads and keeps feedback capture consistent.
—Creates clean, approved CSV exports that can feed analysis and roadmap workflows.

How it works

Listens for a chosen Slack emoji reaction on messages.
Pulls message context, author, channel, permalink, and thread details.
Classifies the feedback and stores raw + structured data in PostgreSQL.
Supports review status updates through API endpoints.
Exports approved and non-exported feedback rows as CSV.

What surprised me

Slack threads are what keep the context window tight, and that assumes people actually use one topic per thread. Next up: channel-wide context for search-style agents when we need it.