mx-validate

An AI review step for mutation testing, a practice that changes source code in small ways to check whether tests detect the changes.

In plain words
What is it for?
Use it to inspect flagged mutants, compare their patches with the current source, and decide whether each mutation is valid or stale.
Why use it?
It determines whether mutants marked for review are still meaningful after the surrounding source code has changed.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/claritune/mutantx/mx-validate
Any agent
npx skills add Claritune/mutantx --skill mx-validate
Clone the repo
git clone --depth 1 https://github.com/Claritune/mutantx

Made for: Claude Code, Codex.

Per session 21 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 649 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00021 $0.00649
Opus 5 $0.00010 $0.00324
Sonnet 5 $0.00004 $0.00130
Haiku 4.5 $0.00002 $0.00065

Measured 2d ago against content hash 6419dc1c74cf, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

mx-validate scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/mx-validate/SKILL.md · 68 lines

How it starts

The opening of the file, as written. The whole thing — 68 lines — stays where its author put it; the contents beside it link to each section on GitHub.

MutantX — Mutation Validation (AI Review)

You are the validation phase of a mutation testing engine. Your job is to review mutants that were flagged as needs-review by the deterministic validator.

These are mutants whose source file changed since generation, but whose patch still applies cleanly. The patch is syntactically applicable, but you need to determine if the mutation is still semantically meaningful.

When This Runs

This skill is invoked after mx-validate.py has already run and marked some mutants as needs-review. If no mutants have that status, tell the user there's nothing to review and exit.

Workflow

1. Read the manifest

Read .mutants/manifest.json. Find all mutants with status: "needs-review".

If there are none, tell the user: "No mutants need review. All are either valid or stale."

2. For each needs-review mutant

Read the mutant's patch file from .mutants/patches/NNN.patch.

Read the current version of the source file the mutant targets.

Evaluate:

  1. Does the mutation still target the same logical code? The patch applies, but did the surrounding code change in a way that shifts what the mutation actually does? For example, if the mutation changes a return value but a new early return was added above it, the mutated line may now be unreachable.

  2. Is the mutation still semantically different from the original? Could the file change have made this mutation equivalent (no observable behavior difference)?

  3. Is the mutation still a realistic bug? If the code around it changed significantly, the mutation might no longer represent a plausible developer mistake.

3. Decide: promote or discard

For each mutant, set the status:

  • pending — the mutation is still valid. The file changed but the mutation still targets the same logic, is semantically different, and represents a realistic bug. Update validation_note to briefly explain why it's still valid.

  • stale — the mutation is no longer meaningful. Update validation_note to explain why (e.g., "Mutated line is now unreachable due to new guard clause at line 42", "Mutation is now equivalent — return value is immediately overwritten").

Read the full file on GitHub · 68 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2d ago First seen · 68 lines · 21 tokens per session scan A 6419dc1c74cf

Subscribe to this mod's changes

mx-validate is a skill published in the GitHub repository Claritune/mutantx (2 stars, last pushed 1mo ago), licensed MIT. It adds 21 tokens to every session and 649 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.