Stryker MCP server: community, described in German

A third-party mutation-testing connector that works, has no license, and hands an English-speaking agent eleven German tool descriptions.

Last verified 25 August 2026 · from The `stryker-mcp-reporter` package, its repository and the server's own tools/list response

This page is one of 90 in a directory of monitoring and developer tools. Each page checks what an AI coding agent can get out of the tool through its MCP server, the connector that lets the agent query the tool directly.

What is the Stryker Mutator MCP server?

The unaffiliated kluth/stryker-mcp-reporter package exposes mutation-test reports from Stryker. Mutation testing changes your source code on purpose, runs the tests, and checks that a test fails. A change no test catches is a survived mutant. The package uses the Model Context Protocol.

The tool set goes beyond running the tool. get_survived_mutants and get_mutant_context reach the mutants your tests missed, which is the output that matters, and suggest_mutant_fixes proposes assertions and boundary tests for them.

Two tools are predictive rather than factual. predict_mutation_impact claims to forecast survivor risk from changed files in under a second, and detect_equivalent_mutants tries to identify mutants no test could ever kill.

Stryker mutation results are exposed only through an unaffiliated npm package, not a server the project publishes. That route covers code and tests only. It holds no CI record, no runtime data, no agent traces and no spend.

This connector answers one question: are the tests testing anything. It also names the assertions that are missing. It says nothing about the build or a running system, so it complements rather than replaces those connections.

What can an AI agent do with Stryker Mutator?

run_targeted_mutation_tests is the one to reach for in a working session. A full mutation run over a real project is long enough that an agent will usually time out waiting for it.

What connecting costs before the first question

A context window is the amount of text a model can hold at once, and connecting uses some of it before the agent asks anything. Every tool the server registers sends its name, description and input schema to the model when the session starts. The tool list is charged whether or not a single tool is called.

11 tools cost between 1,096 and 2,405 tokens, or 0.5% to 1.2% of a 200,000-token window.

It is cheap because it is terse. Descriptions average 97 characters and schemas 242, which is among the leanest measured here.

Encodings: GPT-5.6, GPT-5.5 o200k_base; GPT-4, GPT-3.5 cl100k_base; Claude Haiku 4.5 to Opus 5 count_tokens. Counted 25 August 2026 from tools/list on the released npm package 1.16.1, run locally.

Registration is static, so no project was analyzed during the capture. Figures cover name, description and input schema.

The two models in column one share the o200k_base encoding, so they get one count. The spread is formatting, about 44%. Reasoning tokens are output, spent per request. The tool list is input, charged once at connection.

The Claude figures are measured rather than encoded. Anthropic publishes no local tokenizer. So the definitions were sent to the API's count_tokens endpoint, and a no-tools baseline was subtracted. The range covers two tokenizer generations rather than two serializations. Haiku 4.5 and Opus 4.6 sit at the low end. Opus 5 and Sonnet 5 sit at the high end. The gap between them is about a tenth.

Stryker Mutator MCP server limits: where answers come back incomplete

The first limit decides most evaluations of this connector. The tool descriptions are in a language the prompt is not.

How to configure Stryker Mutator MCP for agents

1. 1

   Read the license question first

   GitHub reports no license for the repository, so there is no explicit grant of rights to use or redistribute it.

2. 2

   Expect German

   All eleven descriptions are in German, which affects how reliably an English-prompted model picks the right tool.

3. 3

   Prefer targeted runs

   A full mutation run is slow, and run_targeted_mutation_tests scopes to files changed in git.

4. 4

   Know which outputs are predictions

   predict_mutation_impact and detect_equivalent_mutants are analyses rather than measurements.

Paste this into your agent’s instructions

Two tools predict rather than measure, and the descriptions are in a different language from the prompt. Both make the wrong tool easy to pick.

When you answer using Stryker MCP tools, state:
- The tool you called and the files it covered.
- Whether a number came from a run or from
  predict_mutation_impact, which forecasts rather than measures.
- Whether a mutant was reported as equivalent, since an
  equivalent mutant cannot be killed by any test.
Never state that data does not exist. State that the run
above produced no survived mutants.
Write your answer in ASD-STE100 Simplified Technical
English. Use short sentences with one idea in each.

Do you need the MCP server at all?

Not in a pipeline. The Stryker CLI is the same runner, and stryker run in a terminal produces the same reports without a connector.

The connector earns its place on the survivor loop: reading get_survived_mutants, pulling get_mutant_context and drafting the missing assertion is exactly what an agent is good at.

For coverage rather than mutation testing, SonarQube and Codacy report it, and BuildPulse tracks it across CI runs.

Connect it when an agent is writing tests. Use the CLI in pipelines, where no model is involved.

Stryker Mutator MCP server setup

The package runs with npx in a project that already has Stryker configured. It needs no credentials, because everything happens locally.

There is no vendor account and no endpoint: the server drives the local Stryker installation and reads its reports.

claude mcp add-json "stryker" '{"command":"npx",
  "args":["-y","stryker-mcp-reporter"]}'

{"mcpServers":{"stryker":{"command":"npx",
  "args":["-y","stryker-mcp-reporter"]}}}

[mcp_servers.stryker]
command = "npx"
args = ["-y", "stryker-mcp-reporter"]

Every other client

Each block below is the configuration for one client, with the file path and the key that client expects.

{"mcpServers":{"stryker":{"command":"npx",
  "args":["-y","stryker-mcp-reporter@1.16.1"]}}}

Read-only access, permissions and security

• There is no license

  GitHub reports none for the repository, which means no explicit permission to use it in a commercial project.

• It runs your test suite

  Mutation testing executes your tests many times over, so an agent calling it starts real local work.

• It writes to the project

  Mutation runs produce reports on disk, and configure_desktop_notifications changes local settings.

• Nobody at Stryker has reviewed it

  The stryker-mutator organization carries no MCP server, and this package is unaffiliated with 0 stars.

Troubleshooting

The agent picks the wrong tool

All eleven descriptions are in German. Name the exact tool in your prompt rather than describing what you want in English.

A run never finishes

Full mutation runs are slow. Use run_targeted_mutation_tests, which scopes to files changed in git.

The score is good but bugs still slip through

Read get_survived_mutants. The score is a summary; the survivors are the specific assertions your tests are missing.

A mutant cannot be killed

It may be equivalent. detect_equivalent_mutants identifies mutants that no test could distinguish from the original.

Stryker Mutator MCP server: Reference

What engineers report

There is no adoption to report and no official alternative, which together are the finding.

The stryker-mutator GitHub organization and the project's site sitemap were searched on 24 August 2026 and carry no MCP server. Tool counts, description lengths and the description language come from the released npm package's own tools/list response.

Should you connect an agent to Stryker Mutator?

Only if a community server is acceptable to you. Best for targeted runs on changed files, with the exact tool named in the prompt.

Strongest fit

TypeScript teams already running Stryker who want an agent to act on survived mutants.

Main advantage

It reaches the output that matters, the survivors, and drafts the assertions that would kill them.

Main weakness

Every tool description is in German, which makes tool selection less reliable for an English-prompted agent.

Operational risk

No license, 0 stars and no review by the Stryker project, on software that runs your test suite.

Frequently asked questions