# chaxiu/munk-ai

Self-improving AI testing engine across Android, iOS, and Web.

Repository: https://github.com/chaxiu/munk-ai
Canonical: https://ross.abutalabs.com/products/munk-ai
Homepage: https://www.munk.sh/
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-08-19T04:59:47+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 98, release rhythm 98, longevity 7
- inputs: {"age_days": 98, "days_push": 14, "days_rel": 14, "gap_med": 16, "n_releases_24m": 2}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1081, forks 102 (observed 2026-08-28T04:03:30.685905+00:00)

## What it is
Munk AI (the open-source Munk Test CLI) is a local-first, self-improving AI testing engine that turns natural-language intent into product-level validation on real Android, iOS, and Web targets. It combines visual-first analysis, real-device execution, structured evidence (screenshots, UI trees, runtime logs), and a knowledge-accumulation loop, exposed via CLI, MCP, local API, and a local Web UI.

## Use cases
- automate mobile app testing on android and ios without writing xpath selectors
- verify ui changes on real devices from natural-language test intent
- set up an ai-powered testing gateway for coding agents
- run visual-first regression tests for web apps
- close the loop between code generation and device-level verification
- collect screenshots, ui trees, and runtime logs as structured test evidence
- improve test coverage over time from accumulated execution evidence

## When to choose
- You need real-device verification (Android, iOS, Web) instead of mocks, static analysis, or fragile selector-based tests
- You want natural-language product requirements turned into automated validation for developers, QA, or coding agents
- You want a local-first, privacy-conscious testing runtime that integrates via CLI, MCP, local API, or Web UI and supports CI regression

## When to avoid
- You need a traditional script-based test framework with fine-grained selector or assertion control
- You need iOS device testing on Windows or Linux, which lack the iOS device bridge
- You require a cloud-hosted device farm or grid at scale rather than a local-first runtime
- You cannot configure the multimodal model backend the engine depends on

## Facets
- artifact type: cli-tool
- maturity: active
- function: e2e-testing, testing, computer-vision, mcp
- domain: testing, developer-tools
- platform: windows, python, cli
- tags: ai-testing, visual-testing, real-device-testing, self-improving, test-automation, qa, mcp-server, local-first, coding-agents, mobile-testing, web-testing, verification-loop, no-xpath, structured-evidence, automation, ai-agents, macos, linux, android, ios

## Member repositories
- chaxiu/munk-ai (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:30.685905+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:51:20.973931+00:00, confidence not recorded.
  - readme: https://github.com/chaxiu/munk-ai (fetched 2026-08-28T04:03:30.685905+00:00, sha 29452a737089)
  - homepage: https://www.munk.sh/ (fetched 2026-08-29T12:53:54.802071+00:00, sha 3b9f437be045)
  - site_page: https://www.munk.sh/docs (fetched 2026-08-29T12:53:54.811330+00:00, sha ef22646b2657)
  - site_page: https://www.munk.sh/install (fetched 2026-08-29T12:53:54.813012+00:00, sha a76760bb894d)
  - site_page: https://www.munk.sh/docs/ai (fetched 2026-08-29T12:53:54.814683+00:00, sha cd8556864f12)
- Data as of 2026-08-30T08:39:29.467469+00:00.