Open-source issue intelligence

Put issue triage on rails.

Tixyo is an open-source AI GitHub issue triage CLI. It turns noisy issues into structured, reviewable decisions—classifying, prioritizing, sizing, routing, and flagging risk while confidence decides what moves.

Dry-run by defaultGitHub nativePython 3.11+
tixyo / triage model ready
$ tixyo triage 482
#482opened 8m ago

Build cache intermittently misses on arm64 runners

CIarm64regression
T Type Bug 96%
C Components Build · CI 91%
P Priority High 88%
A Assignee Needs review 52%
Classify✦Prioritize✦Route✦De-risk✦ Classify✦Prioritize✦Route✦De-risk✦
The workflow

From issue opened
to decision ready.

One command. A bounded context. A decision your team can inspect before it lands.

01
GH
Issue + discussionBody and recent comments
→

Read the whole signal

Tixyo gathers the issue body and recent discussion into a focused context window.

03
Type: Bugready
Owner: Platformreview

Preview, then apply

Dry-run is the default. Explicitly apply only the changes that pass your policy.

Automation with judgment

Fast where it’s safe.
Quiet where it isn’t.

Tixyo separates a model recommendation from permission to mutate your repository. Security-sensitive signals stop normal routing, and low-confidence choices wait for a human.

Read the safety model →
policy.runevaluating
Type: Bug0.96pass
Priority: High0.88pass
Owner: Platform0.52hold
if probability ≥ threshold then eligible
✓
Safe to previewNo repository changes made
One shared language

Your taxonomy,
consistent everywhere.

Define descriptors once. Keep labels aligned across projects. Tixyo understands that an issue has one type, but can touch many components.

tixyo descriptors view --diff tixyo descriptors apply
Type choose one
BugFeatureMaintenance
Component choose many
Source CodeBuildUnit TestsDocumentation
Signals independent
SecurityNeeds infoReproduction
Open source · Python 3.11+

Give your issue queue
a calmer operating system.

Install Tixyo, connect a repository, and preview your first triage decision.

$pip install tixyo
Frequently asked

What teams ask
before they automate.

Short answers about how Tixyo approaches GitHub issue classification and routing.

What is AI issue triage?

AI issue triage analyzes an issue’s title, description, and discussion to recommend structured metadata such as type, priority, size, components, risk signals, and ownership.

Does Tixyo change GitHub issues automatically?

Not by default. Tixyo starts in dry-run mode, displays each recommendation and confidence score, and only writes eligible changes when you explicitly use --apply.

Can an issue have multiple components?

Yes. Tixyo treats component classification as multi-select, so one issue can affect areas such as source code, builds, tests, and documentation at the same time.

How do I install the Tixyo CLI?

Install it from PyPI with pip install tixyo, provide a GitHub token and repository, then run tixyo triage ISSUE_NUMBER to preview a decision.