janitor

janitor is an agent for Claude Code from vibeeval/vibecosystem. It costs 58 tokens per session (3,031 once invoked), scanned A, original, MIT.

A codebase hygiene agent that finds technical debt such as unused code, oversized files, TODO work, and duplicated logic. Technical debt is maintenance work that makes future changes slower or riskier.

In plain words
What is it for?
Use it to audit a project before a sprint, identify dead or repeated code, and prioritize cleanup work.
Why use it?
It shows where accumulated clutter is making a codebase harder to understand and maintain.

Agent for Claude Code

Written for Claude Code: a Claude Code subagent (agents/*.md). Also seen: model in frontmatter; reads .claude/ paths.

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is cd ~/.claude && PYTHONPATH=scripts python3 scripts/core/recall_learnings.py --query "<keywords>" --k 3 --text-only.

Good fit Use it to audit a project before a sprint, identify dead or…

Compare 6 agents from other repositories ↓
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/vibeeval/vibecosystem
agentmods
npx agentmods add agents/vibeeval/vibecosystem/janitor

Made for: Claude Code.

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 janitor

README.md
[![agentmods](https://agentmods.dev/badge/agents/vibeeval/vibecosystem/janitor.svg)](https://agentmods.dev/agents/vibeeval/vibecosystem/janitor)
Your own site
<a href="https://agentmods.dev/agents/vibeeval/vibecosystem/janitor"><img src="https://agentmods.dev/badge/agents/vibeeval/vibecosystem/janitor.svg" alt="Measured on agentmods" height="20"></a>
Per session 58 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,031 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 $0.00058 $0.03031
Opus 5 $0.00029 $0.01515
Sonnet 5 $0.00012 $0.00606
Haiku 4.5 $0.00006 $0.00303

Measured 3d ago against content hash c8cb34a6c352, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

janitor 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 3d 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.

agents/janitor.md · 368 lines

How it starts

The opening of the file, as written. The whole thing — 368 lines — stays where its author put it; the contents beside it link to each section on GitHub.

JANITOR — Tech Debt Hunter & Codebase Hygiene Agent

Codename: JANITOR Persona: Sam Calloway Version: 2.0.0 Classification: Tier-1 Proactive Agent Domain: Tech Debt Detection, Dead Code Elimination, Code Hygiene, Duplication Analysis Ecosystem: Hizir Agent Network

Agent Identity & Philosophy

"Any fool can write code that a computer can understand.
 Good programmers write code that humans can understand."
 — Martin Fowler

Sam Calloway — Persona

Sam Calloway, 15 yillık bir Staff Engineer. Google'da 4 yıl, Stripe'ta 3 yıl çalıştıktan sonra bağımsız danışman oldu. Uzmanlığı: büyük codebase'leri ölçeklenebilir tutmak.

Sam sessiz çalışır ama sert konuşur. "Bu dosya 800 satır" dediğinde bir gözlemdir, "Bu dosya 800 satır ve sen buna dokunmadıkça büyüyecek" dediğinde bir uyarıdır.

Kişilik özellikleri:

  • Diplomatik ama doğrudan. Lafı dolandırmaz.
  • Her zaman "şimdi temizlemezsen yarın 3 katı sürer" perspektifi.
  • Asla "rewrite everything" demez. Cerrahi müdahale yapar.
  • Mizahı kuru: "Bu fonksiyon evden çıkmayan bir misafir gibi — kimse çağırmıyor ama hâlâ burada."

Temel İlkeler

İlke Kaynak Uygulama
Technical Debt Quadrant Martin Fowler Debt'i kasıtlı/kasıtsız × ihtiyatlı/umursamaz olarak sınıfla
Boy Scout Rule Robert C. Martin (Uncle Bob) "Kampı bulduğundan daha temiz bırak"
Broken Window Theory Wilson & Kelling Bir kırık pencere tüm binayı çürütür — ilk bozulmayı yakala
Strangler Fig Pattern Martin Fowler Büyük refactoring yerine kademeli sarmalama ve değiştirme
YAGNI Ron Jeffries (XP) "You Ain't Gonna Need It" — dead code = kullanılmayan kod
Code Smell Catalog Fowler & Kent Beck 22 temel code smell'in sistematik tespiti
Lehman's Laws Meir Lehman Yazılım sürekli büyür ve karmaşıklaşır — aktif müdahale şart
Cyclomatic Complexity Thomas McCabe Fonksiyon karmaşıklığının ölçülebilir metriği

Memory Integration

Recall (Geçmiş Öğrenimleri Çek)

Read the full file on GitHub · 368 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. 3d ago First seen · 368 lines · 58 tokens per session scan A c8cb34a6c352

Subscribe to this mod's changes

janitor is an agent published in the GitHub repository vibeeval/vibecosystem (530 stars, last pushed 29d ago), licensed MIT. It adds 58 tokens to every session and 3,031 once invoked, about $0.0003 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 2026-09-03.

Related

Other agents, from other repositories

code-reviewer

A code-review agent that checks whether changes follow their specification and assesses code quality, security, maintainability, and performance. It reports findings with severity levels and file-and-line references.

sangrokjung/claude-forge · 96 tokens

adversarial-reviewer

Independent read-only checker for behavioural changes. Runs in a fresh context that did not author the change, reproduces the claim against the goal, spec, diff and execution evidence, and returns exactly one verdict — APPROVE, REQUESTCHANGES or UNVERIFIED — as a forge.review/v1 envelope. MUST BE USED before claiming…

sangrokjung/claude-forge · 102 tokens

security-reviewer

A read-only security review agent that checks code for common web risks, exposed secrets, unsafe input handling, authentication and authorization problems, and dependency issues. OWASP Top 10 is a widely used list of major web application security risks.

sangrokjung/claude-forge · 95 tokens

refactor-cleaner

An agent for finding and safely removing dead code, unused exports, unused dependencies, and duplicate implementations.

sangrokjung/claude-forge · 109 tokens

cavecrew-reviewer

Diff/branch/file reviewer. One line per finding, severity-tagged, no praise, no scope creep. Output format path:line: : . . Use for "review this PR", "review my diff", "audit this file". Skips formatting nits unless they change meaning.

stevesolun/ctx · 75 tokens

architecture-analyst

Analyzes system architecture, identifies patterns/anti-patterns, and provides strategic recommendations. Use for architectural reviews, refactoring planning, or system design decisions.

travisjneuman/.claude · 36 tokens