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.
npx agentmods add skills/andrewcigan/vibe-dev-plugin/stucknpx skills add andrewcigan/vibe-dev-plugin --skill stuckgit clone --depth 1 https://github.com/andrewcigan/vibe-dev-pluginWrote 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/skills/andrewcigan/vibe-dev-plugin/stuck)<a href="https://agentmods.dev/skills/andrewcigan/vibe-dev-plugin/stuck"><img src="https://agentmods.dev/badge/skills/andrewcigan/vibe-dev-plugin/stuck.svg" alt="Measured on agentmods" 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 | $0.00068 | $0.01948 |
| Opus 5 | $0.00034 | $0.00974 |
| Sonnet 5 | $0.00014 | $0.00390 |
| Haiku 4.5 | $0.00007 | $0.00195 |
Grade A, and why
stuck 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 192 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/stuck
Формализация правила: «3 попытки — стоп, не 4-я попытка того же».
Триггеры (auto или manual)
Auto:
- 45 мин без commit/test-pass (из tools-allowlist.yaml stuck_detector)
- 3 неуспешных /verify на одну фичу
- 2 одинаковых ошибки подряд (recurrence detect в error-journal)
- stage-verifier вернул FAIL 3 раза
Manual:
- Пользователь сказал «в тупике / крутимся / не получается»
Шаг 0: Эскалация тира ПЕРЕД кворумом (pilotfish N=2)
Прежде чем собирать дорогой LLM-кворум — различи «модель не тянет» и «подход неверен»:
- Если работа/фича шла на дешёвом тире (Sonnet-исполнитель,
implementer) и провалилась 2 раза — первая эскалация НЕ кворум, а поднять тир: перезапусти задачу на Opus или возьми её в main-session (takeover). Запрещён 3-й ретрай того же тира — это и есть залипание. - Кворум и statement (Шаги 1–6) запускаются, ТОЛЬКО если и высший тир (Opus) провалил 2 раза — тогда причина не в модели, а в подходе (нужны принципиально другие гипотезы).
Порядок эскалации: Sonnet ×2 → Opus ×2 → кворум. «Додумал размытый план» на дешёвом тире лечится сменой тира, а не кворумом — это дешевле и часто решает.
Шаг 1: STOP
Немедленно остановить попытки того же подхода. Не делать 4-ю попытку.
Зафиксировать в SESSION.md:
## Stuck Event YYYY-MM-DD HH:MM
- Active feature: feat-XXX
- Attempts made: 3 (или N)
- Last approach: <описание>
- Trigger: auto / manual
Шаг 2: Создать stuck-statement
Subagent stuck-protocol-handler (Sonnet) генерирует docs/stuck-statements/stuck-YYYY-MM-DD-HHMM.md:
# Stuck Statement — <короткое название>
**Дата**: ISO
**Этап pipeline**: <номер>
**Задача**: <что пытались сделать>
## Цель
[Что должны достичь — метрика / проверяемая формулировка]
## Что делали (все попытки)
### Попытка 1
- Подход: ...
- Результат: ...
- Почему не сработало: ...
### Попытка 2-3
...
## Анализ
### Bottleneck (TOC)
[Где именно застряли — конкретный компонент]
### 5 Whys
1-5. Корневая причина: ...
### Диаграмма Ишикавы (опц.)
Man / Method / Machine / Material / Measurement / Environment — что причина?
## Проверенные и отвергнутые гипотезы
- [гипотеза] — отвергнута потому что ...
## Контекст
- Стек, ограничения, данные
- Эталон если есть
## Открытые вопросы
- Что не знаем
- Что нужно уточнить у пользователя
## Прошу у других LLM
Предложите 3-5 новых подходов, принципиально отличных от перечисленных.
Используйте: Lean (муда, poka-yoke, pull), TOC (настоящий bottleneck),
смена абстракции, альтернативные технологии.
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.
- 4d ago First seen · 192 lines · 68 tokens per session scan A 1c7056b1af06
stuck is a skill published in the GitHub repository andrewcigan/vibe-dev-plugin (5 stars, last pushed 1mo ago), licensed MIT. It adds 68 tokens to every session and 1,948 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 2026-08-31.
Other skills, from other repositories
harness
하네스를 구성합니다. 전문 에이전트를 정의하며, 해당 에이전트가 사용할 스킬을 생성하는 메타 스킬. (1) '하네스 구성해줘', '하네스 구축해줘' 요청 시, (2) '하네스 설계', '하네스 엔지니어링' 요청 시, (3) 새로운 도메인/프로젝트에 대한 하네스 기반 자동화 체계를 구축할 때, (4) 하네스 구성을 재구성하거나 확장할 때, (5) '하네스 점검', '하네스 감사', '하네스 현황', '에이전트/스킬 동기화' 등 기존 하네스 운영/유지보수 요청 시 사용.
compose
The mumei orchestrator. For new features, presents a vehicle picker — spec (full SDD workflow: clarification → requirements → design → tasks each auto-reviewed up to 3 iterations → single user approval → Wave-by-Wave implementation → 4-stage review) or plan (Claude Code plan-mode wrapper: hand off to plan mode…
peruse
Plan-vehicle review pipeline. Runs Stage 0 detector (semgrep + osv-scanner) plus security-reviewer and adversarial-reviewer in parallel against the current diff, validates each finding via issue-validator, aggregates a verdict, and writes a review JSON to .mumei/plans/ /reviews/ .json. Triggers when the user invokes…
release
Release a new version of the mumei repository. Invoke when the user gives an explicit release instruction ("release it", "/release", "patch release", "ship 0.2.0"). Takes no argument or "patch" / "minor" / "major" for a SemVer bump, or a direct version such as "0.2.0". Wraps any uncommitted changes into a single…
glean
This skill should be used BEFORE any feature design. It runs structured gleaning with the user — asking 5 high-leverage questions per round, up to 3 rounds, to extract Goal / Scope / Constraints / Edges / Done. Output is saved to .mumei/scratch/ .md and used as input for /mumei:compose. Triggers include "I want to add…
cc-whatsnew
Use when a new Claude Code version is published, when checking the changelog, or when asked "what was added recently?", "what's new in v2.1.x?", or "fetch the official new features". Reports from the official release notes / changelog in two views - an exhaustive feature summary and an applicability verdict for mumei.…