a-check-that-can-only-confirm

a-check-that-can-only-confirm is a skill for Claude Code from synaptiai/synapti-marketplace. It costs 161 tokens per session (910 once invoked), scanned A, original, Apache-2.0.

A verification method for checks such as tests, guards, lint rules, and scans. It requires each check to be shown to detect bad input, ignore correct input, and actually examine its input.

In plain words
What is it for?
It helps verify tests and other automated checks by running cases that must fail, must pass, and must confirm that input was examined.
Why use it?
A green result can be misleading when a check examines nothing or misses the case it was meant to catch. These required test changes expose that problem before the check is shipped.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the flow plugin — 33 skills, 20 commands, 10 agents shipped together

Good fit It helps verify tests and other automated checks by running cases that must fail, must pass, and must confirm that input was examined.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/synaptiai/synapti-marketplace/a-check-that-can-only-confirm
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.

Any agent
npx skills add synaptiai/synapti-marketplace --skill a-check-that-can-only-confirm
Clone the repo
git clone --depth 1 https://github.com/synaptiai/synapti-marketplace

Made for: Claude Code.

Or install flow, the plugin that ships this one along with the rest of its 33 skills, 20 commands, 10 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for a-check-that-can-only-confirm

README.md
[![agentmods](https://agentmods.dev/badge/skills/synaptiai/synapti-marketplace/a-check-that-can-only-confirm/github.svg)](https://agentmods.dev/skills/synaptiai/synapti-marketplace/a-check-that-can-only-confirm)
Your own site
<a href="https://agentmods.dev/skills/synaptiai/synapti-marketplace/a-check-that-can-only-confirm"><img src="https://agentmods.dev/badge/skills/synaptiai/synapti-marketplace/a-check-that-can-only-confirm/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for a-check-that-can-only-confirm

Your own site · 80×15
<a href="https://agentmods.dev/skills/synaptiai/synapti-marketplace/a-check-that-can-only-confirm"><img src="https://agentmods.dev/badge/skills/synaptiai/synapti-marketplace/a-check-that-can-only-confirm.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 161 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 910 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00161 $0.00910
Opus 5.5 $0.00064 $0.00364
Sonnet 5.5 $0.00032 $0.00182
Haiku 4.5 $0.00016 $0.00091

Measured 1mo ago against content hash 2beee0349051, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

a-check-that-can-only-confirm 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 1mo 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.

plugins/flow/skills/learned/a-check-that-can-only-confirm/SKILL.md · 44 lines

How it starts

The opening of the file, as written. The whole thing — 44 lines — stays where its author put it; the contents beside it link to each section on GitHub.

A Check That Can Only Confirm

Contract

Iron law: a check that cannot fail is not a check — prove it can fire, prove it can stay silent, and prove it reached its input. Applies whenever a guard, test, lint rule, roster walk or scan is written or changed. Returns, per check: a must-fire mutant, a must-stay-silent mutant built from a correct input that resembles the incorrect one, an input-removal mutant, an asserted non-zero count of what was examined, and one run against a real full-size input — each one executed, not just written. Permitted skips: none; a check shipped without the three mutants is unverified.

Why a green result proves less than it looks

A guard has two failure modes and only one is loud. Reporting the wrong answer gets noticed. Reaching nothing does not: zero offending files and zero files examined produce the same green. That happens when the matcher is narrower than intended — matching a number only when a unit follows it, handling two of the four spellings a symbol is imported under — so the input the guard exists for never enters it, and nothing in the output says so.

Write three mutants, not one

  • Must-fire — an input the check exists to reject. This is the one everyone writes, and on its own it is satisfied by a matcher that fires on almost anything.
  • Must-stay-silent — a correct input built to resemble the incorrect one as closely as you can. Without it, over-matching stays invisible until the check blocks real work, and the fix under that pressure is to exclude one more case, which just moves the false positive.
  • Input-removal — delete an entry from the roster, empty the directory it walks, rename the symbol it starts from. The check must now fail. This is the mutant that catches a narrowed matcher, because matching nothing and having nothing to match are indistinguishable from the outside.

Make silence say something

Assert and print a count of what the check examined, beside the count of what it found: assert examined > 0 before assert not offending. A run reporting "examined 0 entries, 0 problems" is reporting a defect in itself; without the first number it reads as a pass.

Read the full file on GitHub · 44 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. 1mo ago First seen · 44 lines · 161 tokens per session scan A 2beee0349051

Subscribe to this mod's changes

a-check-that-can-only-confirm is a skill published in the GitHub repository synaptiai/synapti-marketplace (7 stars, last pushed 4d ago), licensed Apache-2.0. It adds 161 tokens to every session and 910 once invoked, about $0.0006 per session on Opus 5.5. 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 2026-09-12.

Related

Other skills, from other repositories

debug

Run /debug to find and fix a bug's root cause: a test failing for an unclear reason, /check verify finding a failure, or behavior being wrong. Runs a reproduce, localize, hypothesize, test, fix, verify loop, makes the minimal fix, and hands a regression test to /test. No features, no extra refactors.

jsmastery-pro/skills · 74 tokens

check

Confirm a change before merge. /check verify drives the real app to prove behavior against the spec (every acceptance criterion met, every surface built). /check review runs a senior code review on a fresh model, one that did not write the code. Verify after /develop, review before a PR. Writes to docs/reviews/, never…

jsmastery-pro/skills · 74 tokens

accessibility-testing

WCAG compliance testing and accessibility quality assurance workflows for iOS apps. Use when validating accessibility labels, testing VoiceOver compatibility, checking contrast ratios, or ensuring WCAG 2.1 compliance. Covers accessibility tree analysis, semantic validation, and automated accessibility testing patterns.

conorluddy/xclaude-plugin · 57 tokens

ios-testing-patterns

XCTest and XCUITest execution workflows and flaky test detection patterns.

conorluddy/xclaude-plugin · 18 tokens

simulator-workflows

Use the executesimulatorcommand MCP tool for all simulator management.

conorluddy/xclaude-plugin · 49 tokens

xcode-workflows

Xcode build system guidance for xcodebuild operations. Use when building iOS projects, running tests, analyzing build failures, or configuring schemes. Covers build/clean/test operations, interpreting xcodebuild output, and troubleshooting common build errors.

conorluddy/xclaude-plugin · 53 tokens