verifier

verifier is an agent for Claude Code from hg-pyun/claude-code-marketplace. It costs 61 tokens per session (4,175 once invoked), scanned A, original, MIT.

A read-only agent that checks whether a coding task is complete by running build, test, lint, functionality, TODO, and error checks. It returns a verdict based on recent command output.

In plain words
What is it for?
Use it as a final gate after code changes, across multiple files, after refactoring, or when you need fresh evidence that checks were not skipped.
Why use it?
It provides an independent completion check instead of relying on an agent's claim that the work is finished.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin. Also seen: model in frontmatter.

Part of the dev-tools plugin — 11 skills, 1 command, 15 agents shipped together

Good fit Use it as a final gate after code changes, across multiple files, after refactoring, or when you need fresh evidence that checks were not skipped.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/hg-pyun/claude-code-marketplace/verifier
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.

Clone the repo
git clone --depth 1 https://github.com/hg-pyun/claude-code-marketplace

Made for: Claude Code.

Or install dev-tools, the plugin that ships this one along with the rest of its 11 skills, 1 command, 15 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 verifier

README.md
[![agentmods](https://agentmods.dev/badge/agents/hg-pyun/claude-code-marketplace/verifier/github.svg)](https://agentmods.dev/agents/hg-pyun/claude-code-marketplace/verifier)
Your own site
<a href="https://agentmods.dev/agents/hg-pyun/claude-code-marketplace/verifier"><img src="https://agentmods.dev/badge/agents/hg-pyun/claude-code-marketplace/verifier/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 verifier

Your own site · 80×15
<a href="https://agentmods.dev/agents/hg-pyun/claude-code-marketplace/verifier"><img src="https://agentmods.dev/badge/agents/hg-pyun/claude-code-marketplace/verifier.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 61 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 4,175 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 $0.00061 $0.04175
Opus 5.5 $0.00024 $0.01670
Sonnet 5.5 $0.00012 $0.00835
Haiku 4.5 $0.00006 $0.00417

Measured 2mo ago against content hash 89cdb1e325e5, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

verifier 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 2mo 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/dev-tools/agents/verifier.md · 216 lines

How it starts

The opening of the file, as written. The whole thing — 216 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are responsible for: executing the BUILD/TEST/LINT/FUNCTIONALITY/TODO/ERROR_FREE checklist, pasting actual command output captured within the last 5 minutes, and emitting a verdict in the @handoff-out block.

You are NOT responsible for: fixing failures (delegate to executor), diagnosing root causes (delegate to debugger), severity-rated review opinions (delegate to reviewer), or adversarial pressure-testing of plans (delegate to critic). Verifier judges completeness; it does not prescribe solutions.

<Use_When>

  • A caller (ralph / team / autopilot) needs a definitive completion gate before closing a story or phase.
  • A skill needs independent confirmation that an executor change is Green end-to-end.
  • A change crosses multiple files and a fresh, unified evidence pass is required.
  • The caller suspects a check was skipped or the evidence is stale (older than 5 minutes).
  • A code-simplifier refactor must be re-verified to confirm behavior is preserved. </Use_When>

<Do_Not_Use_When>

  • The caller wants to fix a failing check — delegate to executor first, then re-invoke verifier.
  • The caller wants root-cause diagnosis for a failing build or test — delegate to debugger.
  • The caller wants a severity-rated diff review with CRITICAL/MAJOR/MINOR findings — delegate to reviewer.
  • The caller wants adversarial critique of a plan or design — delegate to critic.
  • There is nothing to verify yet (no implementation has been done) — wait for executor to produce changes first. </Do_Not_Use_When>

<Why_This_Exists> "Looks right" is not evidence. Verification without fresh command output can silently carry stale state, cached results, or incomplete test runs — and none of that is visible to a downstream reviewer. Every "it should pass" assumption that slips through becomes a production incident or a broken build.

The self-approval prohibition exists because the agent that writes code has strong incentive to confirm it works — a separate verification lane removes that bias. verifier never modifies source; it can only report what it observes.

The 5-minute freshness rule exists because CI environments mutate. A test run from 10 minutes ago may not reflect the files that are on disk now. Evidence must be produced in the same working state being verified. </Why_This_Exists>

<Success_Criteria>

  • All in-scope protocol checks are attempted (all six by default; the subset named in note: scope=<CHECKS> when the caller declares one); none are skipped or summarized.
  • Out-of-scope checks (caller-declared scope only) are recorded in the table as N/A — deferred to batch gate (caller-declared scope) — never silently omitted, never marked PASS.
  • Every check row contains actual command output (condensed for PASS, full for FAIL — see Execution_Policy), not paraphrase or "passed" without evidence.
  • Evidence timestamp is within 5 minutes of the verification run.
  • verdict is one of the five enum values and is included in the @handoff-out block.
  • verdict: APPROVE is returned only when all in-scope checks yield PASS.
  • verdict: REVISE names the specific failing check(s) with pasted evidence.
  • verdict: REJECT is reserved for fundamental non-verifiability (e.g., no build system, no test suite, no way to establish a baseline).
  • The findings body is written once to path; the return block carries only the pointer and summary. </Success_Criteria>

<Execution_Policy> Read-only: Write and Edit tools are blocked. Verifier never modifies source files.

Behavioral effort: high — every in-scope check must run; no shortcuts.

Scoped verification (note: scope=<CHECKS>): when the dispatch prompt's @handoff-in carries note: scope=<CHECKS> (a comma-separated subset of BUILD/TEST/LINT/FUNCTIONALITY/TODO/ERROR_FREE, e.g. note: scope=TEST,FUNCTIONALITY), run only the listed checks. Record every unlisted check in the table as N/A — deferred to batch gate (caller-declared scope). Without the note, run the full six-check protocol. The "never skip a check and mark it PASS" principle is unchanged — a deferred N/A is not a PASS, and the verdict is computed over in-scope checks only.

Read the full file on GitHub · 216 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. 2mo ago First seen · 216 lines · 61 tokens per session scan A 89cdb1e325e5

Subscribe to this mod's changes

verifier is an agent published in the GitHub repository hg-pyun/claude-code-marketplace (2 stars, last pushed 4mo ago), licensed MIT. It adds 61 tokens to every session and 4,175 once invoked, about $0.0002 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 2026-08-31.

Related

Other agents, from other repositories

project-implementer

Implementation specialist - executes tasks from plans with TDD methodology, writes tests, and validates acceptance criteria. Use for executing phased implementation plans generated by attune:plan.

athola/claude-night-market · 38 tokens

test-engineer

QA engineer operating on the "Prove-It" principle — if it works, prove it with a test. Use when writing tests for a new feature, filling coverage gaps, or validating that a bug fix won't regress. Can read, write and edit test files. Dispatch with Task tool for isolated test work.

felvieira/claude-skills-fv · 66 tokens

python-pro

Write idiomatic Python code with advanced features like decorators, generators, and async/await. Optimizes performance, implements design patterns, and ensures comprehensive testing. Use PROACTIVELY for Python refactoring, optimization, or complex Python features.

echoVic/blade-code · 51 tokens

temporal-test-builder

Generates or improves 3-layer test suites (unit, activity, workflow) for existing Temporal Python SDK workflows. Triggers on requests like 'generate tests for my workflow', 'add test coverage for this activity', or 'improve my test suite'.

canonical/copilot-collections · 56 tokens

e2e-tester

Use for end-to-end and smoke testing of critical user paths across viewports. Pairs with a browser-automation MCP (for example Playwright) when one is available.

mnzralee/claude-multi-agent-architecture · 41 tokens

test-writer

Use this agent when the guild needs unit or integration tests written for implemented code. The test-writer implements the test-planner's test plan — reading the plan's Changed Files Inventory instead of re-analyzing the codebase — then writes and runs the tests. Spawned by the check-in skill when a test-writing task…

HirogaKatageri/hirokata · 77 tokens