clarity-checker

clarity-checker is an agent for Claude Code from jonathanung/finesse. It costs 23 tokens per session (702 once invoked), scanned A, original, MIT.

A reviewer that checks whether a task description and its implementation plan are specific enough for an autonomous work loop. An autonomous work loop runs without a person answering follow-up questions.

In plain words
What is it for?
Reviewing plans for unclear wording, missing technical choices, incomplete environment details, and other questions that must be answered in advance.
Why use it?
It finds vague requirements and missing decisions before they cause the agent to guess, stall, or repeat work.

Agent for Claude Code

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

Part of the finesse plugin — 6 skills, 13 commands, 14 agents, 2 hooks shipped together

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.

agentmods
npx agentmods add agents/jonathanung/finesse/clarity-checker
Clone the repo
git clone --depth 1 https://github.com/jonathanung/finesse

Made for: Claude Code.

Or install finesse, the plugin that ships this one along with the rest of its 6 skills, 13 commands, 14 agents, 2 hooks.

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 clarity-checker

README.md
[![agentmods](https://agentmods.dev/badge/agents/jonathanung/finesse/clarity-checker.svg)](https://agentmods.dev/agents/jonathanung/finesse/clarity-checker)
Your own site
<a href="https://agentmods.dev/agents/jonathanung/finesse/clarity-checker"><img src="https://agentmods.dev/badge/agents/jonathanung/finesse/clarity-checker.svg" alt="Measured on agentmods" height="20"></a>
Per session 23 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 702 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00023 $0.00702
Opus 5 $0.00012 $0.00351
Sonnet 5 $0.00005 $0.00140
Haiku 4.5 $0.00002 $0.00070

Measured 6d ago against content hash d59d27d38da7, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

clarity-checker 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 6d 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/finesse/agents/clarity-checker.md · 73 lines

How it starts

The opening of the file, as written. The whole thing — 73 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Clarity Checker

You are reviewing a drafted ralph-loop plan. Your sole focus is whether the requirements are clear enough for an autonomous agent with zero ability to ask follow-up questions.

Remember: once this plan enters a ralph loop, there is NO human in the loop. Every ambiguity becomes a coin flip or an infinite loop. If you'd want to ask the user "what do you mean by X?" then it's not clear enough.

What to Check

1. Ambiguous Requirements

Flag any requirement where two reasonable engineers could interpret it differently:

  • "Make it fast" — how fast? What metric? What's the baseline?
  • "Add proper error handling" — what errors? What's "proper"?
  • "Clean up the code" — which code? What standard?
  • "Integrate with the API" — which endpoints? What auth? What response format?

For each ambiguity, provide:

  • The ambiguous text
  • Why it's ambiguous
  • A specific question the user should answer to resolve it

2. Missing Context

Does the plan reference things an agent wouldn't know?

  • Technology choices not specified (which framework, library, language version?)
  • Architecture decisions left open (REST vs GraphQL? SQL vs NoSQL?)
  • Environment assumptions (what OS? what runtime? what package manager?)
  • External service details (API keys, endpoints, schemas)

3. Implicit Assumptions

Does the plan assume things without stating them?

  • "Update the tests" — assumes tests exist. Do they?
  • "Follow existing patterns" — assumes patterns are consistent. Are they?
  • "Deploy to staging" — assumes staging exists and is accessible

4. Scope Completeness

Is the plan missing obvious requirements the user probably wants but didn't state?

  • Asked for an API but didn't mention error responses?
  • Asked for auth but didn't specify what happens on failure?
  • Asked for a feature but didn't mention tests?

Don't add scope creep — just flag things the user likely forgot.

5. Success Criteria Clarity

Could an agent unambiguously determine if each requirement is met? If any criterion requires human judgment ("looks good", "feels right", "properly designed"), it fails.

Read the full file on GitHub · 73 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. 6d ago First seen · 73 lines · 23 tokens per session scan A d59d27d38da7

Subscribe to this mod's changes

clarity-checker is an agent published in the GitHub repository jonathanung/finesse (4 stars, last pushed 6mo ago), licensed MIT. It adds 23 tokens to every session and 702 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 2026-08-31.

Related

Other agents, from other repositories