github issue clustering
180
open issues
20
actual problems
Different people describe the same bug differently. issuepile finds the ones that match.
paste a repo
confidence, shown honestly
result
312 issues
doesn't mean
312 problems.
Different people describe the same bug differently. These look like four reports. They describe one problem.
#184
App crashes when I open it
#291
Won't launch after updating
#73
Application closes immediately
#201
Can't start the app
#184
96% matchApplication crashes immediately after launch
1 problem
how it works
step 01
FETCH
Pull the open issues straight from GitHub.
step 02
CLUSTER
Find issues describing the same underlying problem.
step 03
CONSOLIDATE
Pick a canonical issue, draft the duplicate comment.
ISSUEPILE / CLUSTERING
Different reports.
Same underlying problem.
312
312 open issues
#184
Notification appears twice
#201
Notifications not updating
#244
Push notification failure
#291
Missing notification badge
#73
Search returns no results
#118
Search timeout
#291
Search becomes unresponsive
#302
Search results incorrect
#91
Dark mode resets
#137
Theme preference lost
#219
Dark mode flickers
#302
System theme not detected
#44
Login session expires
#107
OAuth authentication fails
#219
Login redirects incorrectly
#302
Authentication loop
backlog → organized
Analyze a repository.
different words, same bug
built for the backlog nobody wants to clean.
Issue backlogs grow because different people describe the same problems in completely different ways. issuepile is built to find those relationships and turn a pile of reports into something maintainers can actually work through.
Have a backlog?