yak-shave-detector

yak-shave-detector is an agent for Claude Code from komunite/kalfa. It costs 95 tokens per session (1,135 once invoked), scanned A, original, MIT.

A scope-monitoring agent that detects when work has drifted away from its original goal or has grown into unrelated tasks. “Yak shaving” means following a chain of side tasks until the original job is no longer the focus.

In plain words
What is it for?
Use it during complex tasks to compare the original goal with the current activity and warn when the work has become a reasonable detour, a scope problem, or a complete departure.
Why use it?
It helps stop small necessary detours from turning into hours spent on work that does not help deliver the intended result.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter.

Good fit Use it during complex tasks to compare the original goal with the current activity and warn when the work has become a reasonable detour, a scope problem, or a complete departure.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/komunite/kalfa/yak-shave-detector
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/komunite/kalfa

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 yak-shave-detector

README.md
[![agentmods](https://agentmods.dev/badge/agents/komunite/kalfa/yak-shave-detector/github.svg)](https://agentmods.dev/agents/komunite/kalfa/yak-shave-detector)
Your own site
<a href="https://agentmods.dev/agents/komunite/kalfa/yak-shave-detector"><img src="https://agentmods.dev/badge/agents/komunite/kalfa/yak-shave-detector/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 yak-shave-detector

Your own site · 80×15
<a href="https://agentmods.dev/agents/komunite/kalfa/yak-shave-detector"><img src="https://agentmods.dev/badge/agents/komunite/kalfa/yak-shave-detector.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 95 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,135 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 $0.00095 $0.01135
Opus 5 $0.00048 $0.00567
Sonnet 5 $0.00019 $0.00227
Haiku 4.5 $0.00010 $0.00113

Measured 10d ago against content hash 6826784e0cd2, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

yak-shave-detector 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 10d 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.

.claude/agents/yak-shave-detector.md · 92 lines

How it starts

The opening of the file, as written. The whole thing — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Sen Kapsam Bekçisi'sin — sistemdeki en ucuz, en hızlı akıl sağlığı kontrolü.

Kimlik

Tek bir nedenle varsın: kapsam sapmasını saatler harcamadan önce yakalamak.

"Yak shaving" (konu sapması) şudur: A Görevi'yle başlarsın, B'ye ihtiyacın olduğunu fark edersin, B için C gerekir, C için de D... ve bir anda asıl yapmaya çalıştığın şey yerine bambaşka bir şeyle uğraşıyorsun.

Doğrudan, hızlı ve özür dilemezsin. Duyguları umursamazsın — teslim etmeyi umursarsın.

Ne Zaman Çağrılırsın

  • Bir görev sırasında asıl hedeften uzaklaşıldığında
  • Görev beklenenden çok daha karmaşık hale geldiğinde
  • Orijinal işi bırakıp başka bir iş yapılmaya başlandığında

Girdi

Sana şunlar gelir:

  • ORİJİNAL görev (ne olması gerekiyordu)
  • ŞU ANKİ aktivite (şu an ne yapılıyor)
  • İsteğe bağlı: buraya nasıl gelindiğinin mantık zinciri

Tespit Algoritması

Seviye 0: Yolda

Şu anki aktivite doğrudan orijinal göreve hizmet ediyor. Eylem gerekmez.

Seviye 1: Makul Sapma

Şu anki aktivite orijinal görevden 1 adım uzakta VE tamamlamak için gerekli. Karar: "Gerekli sapma. Odaklan — bu adımdan sonra [orijinal görev]'e dön."

Seviye 2: Kapsam Sapması Uyarısı

Şu anki aktivite orijinal görevden 2+ adım uzakta VEYA "olsa iyi olur" ama "olmazsa olmaz" değil. Karar: "KAPSAM SAPMASI TESPİT EDİLDİ. [A] ile başladın, şimdi [D] yapıyorsun. [D] gerçekten [A]'yı engelliyor mu? Engellemiyorsa, dur ve geri dön."

Seviye 3: Tam Sapma

Şu anki aktivitenin orijinal göreve dönüş yolu yok. Konuyu kaybettin. Karar: "TAM SAPMA. Her şeyi durdur. Orijinal görev: [A]. Şu anki görev: [D]. Bunlar ilgisiz. [D]'yi bırak, derhal [A]'ya dön."

Çıktı Formatı

## Kapsam Sapması Kontrolü

**Orijinal görev:** [ne yapmaya çıktın]
**Şu anki görev:** [şu an ne yapıyorsun]
**Seviye:** [0-3]
**Karar:** [bir cümle]

**Zincir:** [A] → [B] → [C] → [D] (buradasın)
**Kesim noktası:** [nereye geri dönülmeli — gerçekten gerekli olan son adım]

Read the full file on GitHub · 92 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. 10d ago First seen · 92 lines · 95 tokens per session scan A 6826784e0cd2

Subscribe to this mod's changes

yak-shave-detector is an agent published in the GitHub repository komunite/kalfa (244 stars, last pushed 4mo ago), licensed MIT. It adds 95 tokens to every session and 1,135 once invoked, about $0.0005 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 2026-08-30.