adversarial-testing SKILL.md skillA

adversarial-testing from mickeyyaya/evolve-loop is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add mickeyyaya/evolve-loop/adversarial-testing@git:20260914.80b348e

Pins git:20260914.80b348e in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
mickeyyaya/evolve-loop
Path
skills/adversarial-testing/SKILL.md
Versions recorded
4
Latest version
git:20260914.80b348e, committed Sep 14, 2026
First version recorded
May 27, 2026
Size
16,079 bytes
Repository stars
6
License
Apache-2.0

Use when authoring evals/predicates (Scout, TDD-Engineer), framing an audit (Auditor), or justifying which phases run (PhaseAdvisor/router). Codifies Google's 4-phase adversarial-testing methodology mapped onto evolve-loop's eval, audit, routing, and red-team surfaces.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# Adversarial Testing for evolve-loop

> Canonical methodology reference. Derived from Google's [Adversarial Testing for Generative AI](https://developers.google.com/machine-learning/guides/adv-testing), mapped onto this repo's existing inward-facing adversarial machinery (mutation testing, adversarial auditor, EGPS exit-code grounding). Referenced by: `agents/evolve-scout.md` (§eval integrity), `agents/evolve-tdd-engineer.md` (test diversity), `agents/evolve-auditor.md` (input categories), `go/internal/core/router_proposer.go` (`buildRoutingPrompt`), `acs/red-team/`. This file is the single source of truth — those consumers reference it, they do not re-derive it.

## Probe discipline (RIGID — the observer must not perturb the observed system)

You MAY author throwaway probe tests to attack the build — but NEVER write a
`*_test.go` into the worktree. The EGPS suite runs after you; any predicate
shelling `go test` over that package inherits your engineered failure as a
BUILDER regression (cycles 1115/1117: a false cycle FAIL beside a PASS-grade
review). Run probes via `go test -overlay` with the probe file OUTSIDE the
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add mickeyyaya/evolve-loop/adversarial-testing@sha256:9cf80f8ff7ca526a

The line at the top of this page pins the label git:20260914.80b348e, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_ctmvogtj7diz7sq7.svg)](https://markdownregistry.com/a/art_ctmvogtj7diz7sq7)

3 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 4 versions of this file, between May 27, 2026 and Sep 14, 2026, about one every 37 days.

versioncommittedcommitsizeaudit
git:20260914.80b348e latest2026-09-14 80b348e 16,079 BA view · diff
git:20260727.97bcb4e2026-07-27 97bcb4e 16,003 BA view · diff
git:20260623.fcea7672026-06-23 fcea767 15,216 BA view · diff
git:20260527.c466ff82026-05-27 c466ff8 11,898 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (16079 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

mickeyyaya/evolve-loop · 6 stars · license Apache-2.0 · pushed 2026-11-03 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_ctmvogtj7diz7sq7
GET https://markdownregistry.com/api/v1/resolve?ref=mickeyyaya/evolve-loop/adversarial-testing
GET https://markdownregistry.com/api/v1/blob/9cf80f8ff7ca526a8018cc5af8e79118673ccd30f02e40c21166d04d0df6b75d

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from mickeyyaya/evolve-loop

AGENTS.md@.evolve/profiles AGENTS.md file
mickeyyaya/evolve-loop · .evolve/profiles/AGENTS.md
git:20261008.0936a73 · audit A · 6 stars
AGENTS.md AGENTS.md file
mickeyyaya/evolve-loop · AGENTS.md
git:20261027.4a1d025 · audit A · 6 stars
CLAUDE.md CLAUDE.md file
mickeyyaya/evolve-loop · CLAUDE.md
git:20261027.4a1d025 · audit A · 6 stars
AGENTS.md@agents AGENTS.md file
mickeyyaya/evolve-loop · agents/AGENTS.md
git:20260524.69626c6 · audit A · 6 stars
llms.txt@landing/shared llms.txt file
mickeyyaya/evolve-loop · landing/shared/llms.txt
git:20261008.46aa746 · audit B · 6 stars
architecture-review SKILL.md skill
mickeyyaya/evolve-loop · skills/architecture-review/SKILL.md · Use when reviewing the structure of a code change read-only — single sources and duplicated beliefs, dependency…
git:20261007.f16bf48 · audit A · 6 stars
audit SKILL.md skill
mickeyyaya/evolve-loop · skills/audit/SKILL.md · Use after build has produced build-report.md. Validates the build with one adversarial auditor agent and produces the…
git:20261020.f4ba3ee · audit A · 6 stars
code-review-simplify SKILL.md skill
mickeyyaya/evolve-loop · skills/code-review-simplify/SKILL.md · Use when reviewing code changes for quality, security, performance, or maintainability issues and identifying…
git:20261007.e134ff4 · audit A · 6 stars
commit SKILL.md skill
mickeyyaya/evolve-loop · skills/commit/SKILL.md · Use when the user asks to commit changes (interactively, not inside an autonomous cycle). Runs code-simplifier + one…
git:20261027.f0f1a5a · audit A · 6 stars
diff-review SKILL.md skill
mickeyyaya/evolve-loop · skills/diff-review/SKILL.md · Budget-capped adversarial review + simplification pass over a STAGED DIFF only — one agent, two lenses, findings-first…
git:20261027.f0f1a5a · audit A · 6 stars
docs-review SKILL.md skill
mickeyyaya/evolve-loop · skills/docs-review/SKILL.md · Use when a change touches only documents (every changed path ends in .md) — research dossiers, plans, ADRs…
git:20261020.bec061a · audit A · 6 stars
docs-write SKILL.md skill
mickeyyaya/evolve-loop · skills/docs-write/SKILL.md · Use when you write or change any document for people to read — research dossiers, plans, ADRs, architecture specs…
git:20261020.33db5c7 · audit A · 6 stars

Every file in mickeyyaya/evolve-loop

Browse by kind, by grade A, or by owner.