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 instructions/andrewcigan/vibe-dev-plugin/agents-mdgit clone --depth 1 https://github.com/andrewcigan/vibe-dev-pluginWhat 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 | $0.02395 | $0.02395 |
| Opus 5 | $0.01197 | $0.01197 |
| Sonnet 5 | $0.00479 | $0.00479 |
| Haiku 4.5 | $0.00239 | $0.00239 |
Grade A, and why
vibe-dev-plugin AGENTS.md 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 2d 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 — 131 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Vibe Dev v8 Plugin
Это файл для AI-агентов работающих с самим плагином (не с проектами под управлением плагина).
Идентичность плагина
Vibe Dev v8 — harness-first pipeline для разработки продуктов от бизнес-идеи. Главный принцип: «Harness is enforcement, not documentation.» (id плагина — vibe-dev, версия — 8.0.2.)
Структура
vibe-dev/
├── .claude-plugin/
│ ├── plugin.json
│ └── marketplace.json
├── CLAUDE.md ← Мозг плагина (как себя ведут агенты под управлением)
├── README.md / README.ru.md
├── CHANGELOG.md
├── AGENTS.md ← Этот файл (для агентов разрабатывающих сам плагин)
├── agents/ ← 24 агента (реестр роль→модель→тир — docs/agent-registry.md)
│ ├── stage-verifier.md ⭐ v8 — adversarial verifier, disallowedTools=Write/Edit
│ ├── data-model-reviewer.md ⭐ критик модели данных до реализации схемы
│ ├── browser-tester.md ⭐ v7 — Playwright + чтение PNG «глазами»
│ ├── test-researcher.md / user-perspective-critic.md ⭐ dual critique
│ └── ... (полный список — docs/agent-registry.md)
├── skills/ ← 29 команд (skills)
│ ├── new-project/ resume/ feature/ verify/ ship/
│ ├── checkpoint/ ⭐ v8 — управляемая фиксация состояния
│ ├── upgrade-project/ patch-projects/ ⭐ v8 — перевод живых проектов (--soft/--dry-run)
│ ├── doctor/ setup/ hookify/ audit/ stuck/ handoff/ end-session/
│ └── architecture/ research/ dev-plan/ … (полный список — ls skills/)
├── rules/ ← 15 правил и анти-паттернов
│ ├── verification-lanes.md ⭐ evidence по поверхности + logic-lane
│ ├── context-tiers.md ⭐ v8 — трёхуровневая модель контекста
│ ├── model-tier-routing.md ⭐ v8 — тир по стадии + эскалация
│ ├── budgets-and-observations.md / headroom-experiment.md ⭐ v8 (честно discipline)
│ ├── anti-patterns.md ← anti-patterns с ценой из реальных проектов
│ └── quality-gate.md / decision-format.md / lean-toc-language.md
├── templates/ ← 15 шаблонов проектных файлов
│ ├── CLAUDE.md (проектный роутинг ≤200 строк)
│ ├── feature_list.json + feature_list.archive.json ⭐ v8 — архив по ссылке
│ ├── change-proposal.md ⭐ v8 — стадия детализации M/L (P1 US в G/W/T)
│ ├── git-pre-commit.sh ⭐ независимый backstop + 7 блоков гейтов
│ ├── SESSION.md / domain-rules.yaml / error-journal.md / portrait.md
│ └── cold-start.yaml / init.sh / tools-allowlist.yaml / pre-launch-checklist.yaml
├── schemas/ ← feature_list.schema.json (провенанс-голова) + переходы состояний
├── workflow/ ← Методология
│ ├── methodology.md / pipeline.md / stuck-protocol.md
│ └── enforcement-philosophy.md ⭐ 3 честных класса: механизм / подсказка / дисциплина
├── docs/
│ ├── traceability.md ⭐ ЕДИНСТВЕННЫЙ источник числа механизмов (67) и статуса
│ └── agent-registry.md ⭐ v8 — источник истины роль → модель/тир
├── hooks/ ← Pre-action механизмы (авто-загрузка через hooks.json)
│ ├── hooks.json ← авто-регистрация событий при установке
│ ├── dispatch-*.sh ← 6 диспетчеров (pre-tool-use / post / stop / user-prompt / session-start / message-display)
│ ├── pre-compact.sh ⭐ v7 — слепок перед сжатием (страховка)
│ ├── lib/hook-io.sh ← контракт (stdin / permissionDecision / additionalContext) + fail-loud
│ ├── lib/resolve-paths.sh ⭐ v8 — единый резолвер путей харнеса (STRICT/LENIENT)
│ ├── checks/ ← 26 проверок (state-transition, bulk-api, config-protect, secret-scan, …)
│ └── pre-commit-scope.sh ← git-hook, closes WIP=1 enforcement
├── scripts/ ← 16 скриптов
│ ├── check-plugin-self.sh ⭐ self-check (45 разделов) — главный гейт разработки
│ ├── check-traceability.sh / check-no-personal-data.sh
│ ├── record-change.sh ⭐ v8 — единственный crash-safe путь записи провенанса
│ ├── checkpoint.sh / archive-features.sh / migrate-provenance.sh ⭐ v8
│ ├── upgrade-project.sh / patch-projects.sh / install-precommit.sh
│ └── audit-health.sh / journal-audit.sh / stuck-watcher.sh
└── tests/hooks/ ← 37 тестовых наборов + фикстуры (в т.ч. 6 обезличенных боевых)
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.
- 2d ago First seen · 131 lines · 2,395 tokens per session scan A 535312573cdc
vibe-dev-plugin AGENTS.md is an instructions file published in the GitHub repository andrewcigan/vibe-dev-plugin (5 stars, last pushed 1mo ago), licensed MIT. It adds 2,395 tokens to every session, about $0.0120 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 instructions, from other repositories
qm AGENTS.md
AGENTS.md instructions for yc-software/qm, covering qm, working on the code, private forks and durable by default.
vibe-coding-prompt-template backend.instructions.md
Instructions for KhazP/vibe-coding-prompt-template: Read AGENTS.md, agentdocs/techstack.md, and agentdocs/codepatterns.md.
cherry-studio AGENTS.md
AGENTS.md instructions for CherryHQ/cherry-studio, a project described as: AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs.
sprawling AGENTS.md
Instructions for 2youg1/sprawling, covering agents.md — how work is done in this repository, the loop, read before you write, one change, five steps and the rules a machine holds.
easy-vibe AGENTS.md
AGENTS.md instructions for datawhalechina/easy-vibe, covering repository guidelines, project structure & module organization, build, test, and development commands, coding style & naming conventions and testing guidelines.
tree-sitter-analyzer CLAUDE.md
Instructions for aimasteracc/tree-sitter-analyzer, covering ruflo — claude code configuration, rules, test quality rules — locked (2026-06-28), t-1: no new test files for existing plugins (blocker) and t-2: no weak assertions — ratchet-enforced (blocker).