A powerful utility for identifying and prioritizing flaky tests by analyzing JUnit retry artifacts within your CI pipeline.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install ci-flake-triage
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install ci-flake-triage using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
CI Flake Triage is a specialized tool designed for engineering teams looking to improve test suite reliability and developer productivity. By analyzing multiple JUnit XML reports—such as initial test runs and subsequent automated retries—this skill identifies tests that exhibit inconsistent behavior. It provides developers with the clarity needed to distinguish between persistent failures that require immediate code fixes and flaky tests that undermine CI stability.
Integrating this tool into your workflow via Openclaw Skills ensures that your team focuses stabilization efforts on the most impactful issues. By reducing noise in your build reports, you can maintain a higher level of trust in your automated testing suite while streamlining the debugging process for complex distributed systems.
Ensure you have bash and python3 installed on your environment. You can execute the triage script directly by targeting your test artifacts directory. This skill is designed to work seamlessly within the Openclaw Skills ecosystem.
# Basic execution with custom glob
JUNIT_GLOB='artifacts/junit/**/*.xml' \
bash skills/ci-flake-triage/scripts/triage-flakes.sh
# Run with JSON output for automated CI ingestion
OUTPUT_FORMAT=json \
FAIL_ON_PERSISTENT=1 \
bash skills/ci-flake-triage/scripts/triage-flakes.sh
The skill processes standard JUnit XML files and generates a summary based on the following taxonomy:
| Property | Description |
|---|---|
| Test Case Name | The unique identifier for the specific test scenario captured in the XML. |
| Status Aggregation | A summary of all outcomes for a specific test (e.g., 1 pass, 2 failures). |
| Flaky Candidate | Flagged when a test has both fail-like and pass outcomes in the same set. |
| Persistent Failure | Flagged when a test has only failure or error outcomes across all retries. |
| Triage Top | A configurable limit (default 20) on the number of flaky tests displayed in the report. |
Loading
A streamlined MCP-based skill for automating McDonald's China delivery orders, menu lookups, and coupon management via official API protocols.

A hosted OAuth proxy that enables seamless data retrieval from Oura Ring and Google Calendar using a single master API key.

A self-hosted Node.js service that provides a persistent REST API for Google Calendar with automated OAuth token refreshing.

A specialized AI sub-agent for automating outbound property showing calls and managing booking statuses.

Statically audit Dockerfiles to identify and mitigate security risks before container deployment.

A robust auditing tool to detect environment variable alias conflicts and missing required configurations before deployment.








































