Is the Adversarial Verify skill free?

Yes. The skill itself is free from NUX-Design/claude-skills-fable-opus. SkillProof publishes the install command and an independent test verdict at no cost.

What is the Adversarial Verify SkillProof Score?

9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.

How do I install Adversarial Verify?

Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.

Can I use Adversarial Verify with Cursor, Copilot, Gemini CLI, Codex or other AI tools?

The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.

Adversarial Verify

Refute your own work before presenting it: attack inputs, assumptions, evidence

by NUX-Design · NUX-Design/claude-skills-fable-opus

Tested · Works

Install — copy the command Download source ↗

Test report

Verdict: Tested · Works
Score: 9.2/10
Tested: Jul 19, 2026
Environment: Claude Code 2.x (agent harness)
Upstream re-checked: Jul 30, 2026 · bccbfbf

⚠ This skill is no longer available upstream. Our re-check on Aug 10, 2026 couldn't find it any more (repo unreachable/deleted). The test below is what we measured on Jul 19, 2026 and we're leaving it up as a record — but there is nothing left to install, so we've removed the command.

Ran the head-to-head on a second_largest(nums) task: the answer shipped without the refutation pass crashed with IndexError on single-element, all-equal, and empty inputs, while the skill's 'attack the inputs' step surfaced exactly those three cases and produced a version that handled them. The install is a clean prose-only skill with no scripts to break, and the verbatim copy landed SKILL.md from a bare HOME. Trigger wording cleanly separates substantive work from trivial questions. The one soft spot is the README's 12-0-2 blind-benchmark claim, which is not independently reproducible here, but the skill mechanism itself delivered a concrete, measurable catch.

Scored on four weighted criteria — install, triggering, output vs. baseline, docs. How scoring works

Installs cleanly 5/5
Triggers reliably 5/5
Output vs. baseline 8/10
Docs & honesty 5/5

What Adversarial Verify does

A self-review discipline that flips you from author to attacker after finishing code, analysis, or a document and before calling it done. It walks a fixed refutation pass (state the claim, attack requirements/inputs/assumptions/evidence, run the strongest check) and only ships work that survives. Triggers on any substantive piece of work about to be presented as finished.

How to install Adversarial Verify

Nothing to install: the source repository no longer has this skill. If the author brings it back, our daily re-check will pick it up and the command will reappear here.

Commands — how to trigger Adversarial Verify

/adversarial-verify Refute your own work before presenting it: attack inputs, assumptions, evidence

It also activates on plain-language prompts like these:

I just finished this migration script — is it really done?
Here's my analysis, ready to ship — poke holes in it first.
Double-check this config for edge cases before I present it.

Frequently asked questions

Is the Adversarial Verify skill free?: Yes. The skill itself is free from NUX-Design/claude-skills-fable-opus. SkillProof publishes the install command and an independent test verdict at no cost.
Does Adversarial Verify work with Claude Code?: We tested it with Claude Code 2.x (agent harness) on Jul 19, 2026. Verdict: Tested · Works. Ran the head-to-head on a second_largest(nums) task: the answer shipped without the refutation pass crashed with IndexError on single-element, all-equal, and empty inputs, while the skill's 'attack the inputs' step surfaced exactly those three cases and produced a version that handled them. The install is a clean prose-only skill with no scripts to break, and the verbatim copy landed SKILL.md from a bare HOME. Trigger wording cleanly separates substantive work from trivial questions. The one soft spot is the README's 12-0-2 blind-benchmark claim, which is not independently reproducible here, but the skill mechanism itself delivered a concrete, measurable catch.
What is the Adversarial Verify SkillProof Score?: 9.2/10 — installs cleanly 5/5, triggers reliably 5/5, output vs. baseline 8/10, docs & honesty 5/5.
How do I install Adversarial Verify?: Copy the install command from this page, run it in your terminal, and restart Claude Code. Skills live in ~/.claude/skills/ (global) or .claude/skills/ inside a project.
Can I use Adversarial Verify with Cursor, Copilot, Gemini CLI, Codex or other AI tools?: The SKILL.md format is native to Claude (Claude Code, Desktop, claude.ai). The instructions inside adapt to other assistants: Cursor rules, GitHub Copilot instructions, Windsurf rules, Custom GPTs, AGENTS.md for OpenAI Codex, and GEMINI.md for Google Gemini CLI — our conversion guides cover each, and the free converter on the tools page does the wrapping for you.