Skip to content
Get packs

Feedback Sweep

Collects new feedback from Slack/GitHub, analyzes recordings, checks fixes, and builds a plan for lfg

Source Download skill
Repository
EveryInc/compound-engineering-plugin
Stars
25,850
Licence
MIT
Files
14 files

What it does

Once installed, when you run this skill the agent checks the feedback sources you configure (such as Slack channels and GitHub Issues) for new items since the last run. It can acknowledge items at their source, analyze attached screen recordings or audio into transcripts and notes, check whether claimed fixes have been merged to the default branch, and write a rolling plan file you can hand to the lfg workflow. It keeps its own state file so runs don't repeat work.

Great for

  • Product or engineering teams who triage user feedback from Slack or GitHub Issues
  • Maintainers who want recurring feedback sweeps, including scheduled non-interactive runs
  • People already using the compound-engineering lfg workflow who want feedback turned into a plan

Skip it if

  • Teams that do not want an agent posting acknowledgements to Slack or GitHub, since source-side writes happen for approved sources (risk: unwanted public replies)
  • Sensitive feedback environments without reviewing the sensitive flag and state file, since media and item content are stored in the repo-internal state and plan (risk: data exposure)
  • Anyone expecting email to work reliably, since email is marked experimental in the skill description

Try asking

  • “Sweep our configured feedback sources for new items and summarize what needs attention”
  • “Analyze the Riffrec recording attached to the latest feedback and add findings to the plan”
  • “Check whether the fixes claimed in open feedback items have been merged to main”
  • “Reconfigure the feedback sources we sweep”
Step-by-step guide: how to use Feedback SweepInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 14 original files, the MIT licence and an attribution note.

    ce-sweep.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip ce-sweep.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip ce-sweep.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Run the first-time setup interactively so you can review which sources are approved for source-side writes before any acknowledgements are posted.
  • Keep sources set to unapproved until you have checked the plan output, so the agent only fetches and records items.
  • Keep raw recordings out of commits; the skill is designed to commit only the plan and repo-internal state.

What's inside

  • agents/openai.yaml
  • references/agents/media-analyzer.md
  • references/interview.md
  • references/model-tiers.md
  • references/plan-template.md
  • references/run.md
  • references/sources/email.md
  • references/sources/github-issues.md
  • references/sources/slack.md
  • references/state-schema.md
  • references/subagent-template.md
  • scripts/analyze_riffrec_zip.py
  • scripts/sweep-state.py
  • SKILL.md

Redistributed unmodified from EveryInc/compound-engineering-plugin under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.