ariadne

io.github.MikhailHalPublished by GitHub account MikhailHalv0.4.0Updated Oct 24, 2026

Returns only the unit tests affected by your current Kotlin/Android code changes.

VerifiedSTDIODesktop onlyDeveloper ToolsData & Analytics

Overview

AI-generated overview

An MCP server that analyzes Kotlin/Android code changes and returns only the unit tests affected by them.

What you can do
Ariadne exposes a get_affected_tests tool that takes a Kotlin project path and returns a sorted list of affected test fully qualified names. It runs git diff internally against a base branch or HEAD, then uses the sazanami Kotlin Analysis API to build a call graph and find affected tests. It reports explicitly when changes fall outside analyzed sources, such as build scripts or resources, and returns an error on its 120-second analysis timeout.
When to use it
Useful during an agent's edit-verify loop on Kotlin or Android projects, where running only the affected tests gives faster feedback than the full suite. It is not a replacement for CI: the README advises keeping a full or conservative test run before merging.
Requirements
Runs as a local process over stdio. Needs JDK 21 or later and Git for diff detection. Installable via Homebrew, a Docker image, or a release JAR; the Docker route requires mounting the project and passing /workspace as project_path. No accounts, API keys, or environment variables are declared.
Before you install
It reads the project and runs git diff locally; no credentials are requested. Selection is static-analysis based and can miss dependencies hidden by reflection, DI wiring, framework-invoked callbacks, or lifecycle-started subscriptions, and it does not analyze build scripts, resources, or instrumented tests. Keep a full test run in CI before merging.

Installation

In SourceWeft

  1. Open ariadne in the dashboard and add it to a workspace.
  2. Enable the server for the chats that should use its tools.

Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.

Other MCP clients

Follow the launch instructions in the repository.

README

[ariadne banner]

ariadne

MCP Server for Affected Test Selection

[Release] [MCP Registry] [Homebrew] [Container image] [License] [Kotlin]


ariadne is an MCP (Model Context Protocol) server that provides AI agents with the ability to identify affected tests. Powered by sazanami, it analyzes code changes and returns only the tests that need to be run.

[ariadne MCP server]

[!IMPORTANT]

⚠️ MUST READ: Speed over Completeness

ariadne is built for the agent inner loop — edit, verify, commit — where fast feedback matters more than exhaustive selection. Static analysis cannot trace every execution path: reflection, DI frameworks, and data-flow indirection (e.g., Flux/MVI dispatch) can hide dependencies from any affected-test-selection tool, not just ariadne.

Always keep a final line of defense in CI. Run the full test suite (or a conservative selection) before merging. ariadne narrows what an agent runs while iterating; it is not a replacement for CI.

When ariadne detects changes it cannot analyze (build scripts, resources, unscanned source sets), it says so explicitly in the tool response instead of silently reporting "no affected tests".

Features

  • MCP Integration — Works with Claude Code, Claude Desktop, and other MCP-compatible clients
  • Automatic Git Diff — No need to pass diff manually; ariadne runs git diff internally
  • Powered by sazanami — Uses Kotlin Analysis API for accurate static analysis

Installation

Homebrew (recommended)

bash
brew install mikhailhal/tap/ariadne

Then register it with your MCP client — for Claude Code:

bash
claude mcp add ariadne -- ariadne

Or add to your MCP client configuration manually (e.g., Claude Desktop):

json
{  "mcpServers": {    "ariadne": {      "command": "ariadne"    }  }}

Docker

bash
docker run -i --rm -v /path/to/project:/workspace ghcr.io/mikhailhal/ariadne

Mount the project you want analyzed and pass /workspace as project_path. The image is also listed in the official MCP Registry as io.github.MikhailHal/ariadne.

Manual (release JAR)

Download ariadne-<version>-all.jar from Releases (requires JDK 21+) and configure your client with "command": "java", "args": ["-jar", "/path/to/ariadne-<version>-all.jar"].

Build from Source

bash
git clone --recursive https://github.com/MikhailHal/ariadne.gitcd ariadne./gradlew shadowJar   # fat JAR: build/libs/ariadne-<version>-all.jar

Usage

Once configured, AI agents can use the get_affected_tests tool:

Tool: get_affected_tests

Parameters:

  • project_path (required) — Path to the Kotlin project
  • scope (optional, deep | shallow, default deep) — how far back to look:
    • deep — all changes since base_branch (committed and uncommitted). Safest; the whole branch is covered so nothing you already committed slips through unverified. May select more tests.
    • shallow — uncommitted changes only (diff against HEAD). Fastest, for the tight edit loop. Verifying already-committed work is left to the caller. base_branch is ignored.
  • base_branch (optional, deep scope only) — Branch to compare against. When omitted, ariadne uses the repository's default branch (origin/HEAD). If that is not set (e.g. a repo with no remote), it returns an error asking you to pass base_branch explicitly rather than guessing.

Returns:

  • List of affected test FQNs (fully qualified names), sorted
  • If the diff contains changes outside the analyzed Kotlin sources (build scripts, resources, unscanned source sets), a note is appended recommending a full test run for those changes
  • Analysis is bounded by a 120s timeout; on timeout an explicit error is returned

Example

Agent request:

json
{  "name": "get_affected_tests",  "arguments": {    "project_path": "/path/to/your/kotlin/project"  }}

Response:

com.example.UserServiceTest.testCreateUsercom.example.UserRepositoryTest.testSave

How It Works

┌─────────────┐     ┌─────────────┐     ┌─────────────┐│  MCP Client │ ──▶ │   ariadne   │ ──▶ │  sazanami   ││  (Agent)    │     │ (MCP Server)│     │  (Analysis) │└─────────────┘     └─────────────┘     └─────────────┘                           │                           ▼                    ┌─────────────┐                    │  git diff   │                    └─────────────┘
  1. Agent calls tool — Passes project path to ariadne
  2. Run git diff — ariadne executes git diff --unified=0 against base branch
  3. Analyze with sazanami — Build call graph and find affected tests
  4. Return results — List of test FQNs returned to agent

Real-World Validation: Now in Android

Measured against Now in Android (Google's reference Android app — 34 modules, ~268 Kotlin files):

MetricResult
Recall audit — 19 target functions across all layers18/18 valid targets detected (the 19th had no exercising unit test; correctly not selected)
End-to-end response time~4s (module discovery + call-graph build + BFS)
Module discovery34 modules via settings.gradle.kts, incl. nested modules and type-safe accessor dependencies
Source setsmain, debug, prod, benchmark, testDemo, … discovered per module (androidTest* excluded by design)

Verified patterns include repositories behind project interfaces, a library-interface override (androidx.datastore.Serializer), operator fun invoke use cases, @Composable functions, extension mappers, ViewModel property-initializer chains, and callable references. Two representative results:

  • Changing core:common's asResult() selects 14 tests across three modules, including ViewModel tests reachable only through val uiState = ...stateIn(...)
  • Changing the mapper PopulatedNewsResource.asExternalModel() selects 14 tests, including 11 repository tests reachable only through .map(Type::mapper) chains

Test-class selection rate

Every unit-test class in Now in Android was measured by changing a function in the class it tests and checking whether that test class was selected:

Test styleSelected
Plain unit tests (construct the object, call it)13 / 13 valid targets
Robolectric / Compose screenshot tests12 / 12
Framework-dispatched callbacks (lint Detector)0 / 2 — see below

Robolectric turned out not to be a barrier: those tests call the composable themselves (setContent { NiaTheme { ... } }), so the call exists in the source. What decides coverage is not the test runner but whether the test's own code contains the call.

What ariadne cannot see

The rule of thumb: if the framework calls your code instead of your test calling it, ariadne cannot connect them. These are limits of static analysis, not bugs — plan your CI safety net around them:

PatternStatus
Framework-invoked callbacks — Fragment/Activity lifecycle (launchFragmentInContainer), lint Detector methods, Application.onCreateNot traced: no call written in the test
Reflection / DI-container wiringNot traced
UDF dispatch (Flux/MVI)dispatch → collect is never an edge, but wiring in init (or a start() the test calls) is covered conservatively via constructor chains. Subscriptions started by DI/lifecycle are not covered (sazanami#38)
stateIn / shareIn chains (map, onEach, flatMapLatest, combine)Covered — verified with exact selection
Instrumented tests (androidTest*)Out of scope by design
Build scripts, resources, unscanned source setsNot analyzed — reported explicitly in the tool response
Same-name top-level extensions in one packageOver-selected (receiver types are not part of top-level FQNs) — safe direction
KMP source sets (commonMain, expect/actual)Enumerated, but resolution quality unverified (#1)

Full audit notes: sazanami#29, sazanami#38.

Requirements

  • JDK 21 or later
  • Git — For diff detection

Limitations

  • Module discovery is convention-based: it parses settings.gradle(.kts) includes, enumerates src/<sourceSet>/{kotlin,java} layouts, and reads project(":x") / type-safe accessor dependencies from build files. Dynamic includes, projectDir remapping, custom srcDirs, and dependencies injected by convention plugins are not detected — see #1
  • Full graph rebuild on each request (no caching yet); analysis is capped at 120s
  • See What ariadne cannot see for analysis-level gaps

License

Copyright 2025 ariadne contributors
Licensed under the Apache License, Version 2.0 (the "License");you may not use this file except in compliance with the License.You may obtain a copy of the License at
    http://www.apache.org/licenses/LICENSE-2.0

See LICENSE for the full text.

Source: README.md at commit 057144e

Tools

0
SourceWeft has not started this package in its sandbox yet, so its tools are not known.

Version history

1
  1. v0.4.0LatestSep 16, 2026

Score

50% of the applicable checks met

How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.

  • Reached in the last 30 days25 pts

    Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.

  • Auth declared10 pts

    Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.

  • Tools listed15 pts

    Answer tools/list without a credential, or ship a package that starts in a sandbox without one.

  • DescriptionRequired15 pts
  • Logo10 pts
  • Documentation15 pts
  • License10 pts