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.
git clone --depth 1 https://github.com/artemnovichkov/skillsWrote 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.
[](https://agentmods.dev/commands/artemnovichkov/skills/verify)<a href="https://agentmods.dev/commands/artemnovichkov/skills/verify"><img src="https://agentmods.dev/badge/commands/artemnovichkov/skills/verify/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.
<a href="https://agentmods.dev/commands/artemnovichkov/skills/verify"><img src="https://agentmods.dev/badge/commands/artemnovichkov/skills/verify.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00025 | $0.00617 |
| Opus 5 | $0.00013 | $0.00309 |
| Sonnet 5 | $0.00005 | $0.00123 |
| Haiku 4.5 | $0.00003 | $0.00062 |
Grade A, and why
verify 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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Stage 4A — Every session checks its own work
Never hand a change to a human without having checked it yourself. This command runs the loop and reports literal output, not an assessment.
$ARGUMENTS may be --lock-tests / --unlock-tests (see step 5) or empty.
Workflow
1. Find the verification command
In order:
CLAUDE.md— the build/test/lint commands and their healthy output (put there by/sdlc:init)Makefile,package.jsonscripts,Package.swift, project scheme- Ask the user, then offer to write the answer into
CLAUDE.mdso this stage never has to ask again
2. Establish what "done" means
Read sdlc/plan/<slug>.md if one matches the current branch or change, and take the done-condition from its Tests section. Otherwise ask for a quantifiable target: which tests pass, what the build must produce, what the UI must show.
3. Run everything
Run build, tests, and lint — all of them, before reporting anything. Paste the real output, trimmed but not paraphrased. If a command fails, that is the result; report it as a failure with the output rather than describing what you would fix.
For UI changes, capture evidence: a screenshot, a rendered preview, or a driven run of the app. Iterate two or three rounds against the expected appearance before showing it to the user.
4. Delegate the behavioural pass
For anything beyond a unit-test-shaped change, launch the verifier subagent bundled with this skill. It runs the app, exercises the changed behaviour and the flows next to it, and reports what worked without fixing anything. Its findings go to the user as-is.
5. Test-file lock (bug fixes)
For a bug fix the order is: write the failing test first, commit it, then fix the code without touching the test.
--lock-testscreatessdlc/.lock-tests. While that file exists, this skill'sPreToolUsehook blocks edits to test files, so the fix cannot be made to pass by weakening the check.--unlock-testsremoves it. Always remove the lock once the fix is verified, and tell the user it is off.
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.
- yesterday First seen · 48 lines · 25 tokens per session scan A 60fb4400051c
verify is a command published in the GitHub repository artemnovichkov/skills (36 stars, last pushed yesterday), licensed MIT. It adds 25 tokens to every session and 617 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 2026-09-07.
Other commands, from other repositories
test
Detect the project's test framework and run the test suite.
test-prompt
Test an AI prompt against multiple scenarios to verify consistent, quality output.
create-component
Create a frontend component with proper structure, types, and accessibility.
run-e2e
Run end-to-end tests against the application using Playwright or Cypress.
review
Review the current branch's changes for quality, bugs, and best practices.
improve-test-quality
Analyze and improve existing test quality with better assertions, coverage, and maintainability.