ArcReel
ArcReel/ArcReel/AGENTS.md
ArcReel AI 视频创作平台,将小说、剧本或创作构想转化为短视频。三层结构:`frontend/`(React SPA)→ `server/`(FastAPI,`agent_runtime/` 封装 Claude Agent SDK)→ `lib/`(核心库)。内嵌创作 Agent 的配置源在 `agent_runtime_profile/`,与开发态 `.claude/` 分离。uv workspace
How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
ArcReel/ArcReel/AGENTS.md
ArcReel AI 视频创作平台,将小说、剧本或创作构想转化为短视频。三层结构:`frontend/`(React SPA)→ `server/`(FastAPI,`agent_runtime/` 封装 Claude Agent SDK)→ `lib/`(核心库)。内嵌创作 Agent 的配置源在 `agent_runtime_profile/`,与开发态 `.claude/` 分离。uv workspace
agentscope-ai/agentscope-java/AGENTS.md
storage, protocol, channel, sandbox, framework integrations. - `agentscope-service/` — Java services; `service-controlplane/` is Go, `frontend/` is React. - `agentscope-examples/` — runnable examples. - `agentscope-dependencies-bom/` — third-party versions. - `agentscope-distribution/` — published
Steady5 mentions in 2 weeks
No other project uses it yet
baserow/baserow/AGENTS.md
backend/src`, shared backend tests in `backend/tests`, and the main Nuxt app in `web-frontend/` (`modules/`, `server/`, `test/`, `stories/`). Paid extensions mirror that layout in `premium/backend`, `premium/web-frontend`, `enterprise/backend`, and `enterprise/web-frontend
fontsource/fontsource/AGENTS.md
meaningful output contracts: API responses, headers, generated CSS, fonts, and archives. Do not add frontend interaction, reactive-store, internal orchestration, or cache-mechanics tests by default. Zero new tests
bookorbit/bookorbit/AGENTS.md
libraries. Avoid unbounded queries, loading entire libraries into memory, returning full collections to the frontend, N+1 query patterns, or expensive client-side processing over full result sets. - Use user
No other project uses it yet
lightdash/lightdash/AGENTS.md
flag, read [docs/feature-flags.md](docs/feature-flags.md). Use `FeatureFlagModel.get` / `FeatureFlagService.get` in backend services and `useServerFeatureFlag` in the frontend. Do not create an ENV-only or `NODE_ENV` rollout gate: the shared resolver supports
Steady46 mentions in 2 weeks
No other project uses it yet
lightdash/lightdash/docker/AGENTS.md
variables are deduplicated using YAML anchors (`x-lightdash-*`) **Port Allocation**: - 8080: Backend API - 3000: Frontend development server - 9090: Prometheus metrics - 5432: PostgreSQL database - 9000/9001: MinIO storage - 8025: Mailpit
Steady46 mentions in 2 weeks
No other project uses it yet
Akegarasu/lora-scripts/AGENTS.md
docs/agent/dev/README.md);**人类说明**:[文档索引](docs/README.md)。 - 后端检查:`python -m unittest discover -s mikazuki/tests -t .`;前端检查:在 `frontend` 执行 `corepack pnpm test` 和 `corepack pnpm run build`。 - 复用运行中的后端;同一工作区只运行一个队列服务。模型、数据、训练结果和本地配置不提交。 - `scripts/dev
EverMind-AI/Raven/AGENTS.md
trees (`bridge/`, `ui-web/`, `ui-tui/`) are exempt from that extension list: a product frontend carries its own entry HTML and icon SVGs as source, not as report assets. Everything
brookhong/Surfingkeys/AGENTS.md
them starts a fresh pile every time. - Reaching another tab means the background: the frontend iframe cannot address a tab it does not live in, so it is `RUNTIME('getTabMarkdown
tw93/Kaku/kaku-gui/AGENTS.md
confirmed path may auto-confirm. CLI direct and brew paths confirm in `kaku/src/update.rs`. Guards: `frontend::tests::user_facing_update_events_route_through_confirm` and `update::imp::tests::brew_and_direct
ModelEngine-Group/nexent/AGENTS.md
instructions ## Repository map - `backend/` contains the FastAPI HTTP API; `sdk/nexent/` contains the agent framework. - `frontend/` contains the Next.js UI. Deployment resources live in `deploy/`, `docker/`, and `k8s/`. - Python tests live
WeblateOrg/weblate/AGENTS.md
templates, use `{% translate %}` / `{% blocktranslate %}` for translatable text. - Preserve accessibility and the existing Bootstrap/jQuery-based frontend patterns. For user-facing HTML, CSS, or JavaScript changes, follow `ACCESSIBILITY.md` and `docs/contributing/frontend.rst`, including keyboard navigation
shuakami/qq-chat-exporter/AGENTS.md
overlay runtime under plugin `node_modules/NapCatQQ` | overlay generator and NapCat source snapshot | `node plugins/qq-chat-exporter/tools/create-overlay-runtime.cjs` | | Frontend static export under `qce-v4-tool/out/` | `qce-v4-tool/app`, `components`, `hooks
greatfrontend/top-reactjs-interview-questions/AGENTS.md
Vitest, tsdown, Oxlint, Oxfmt, and Vite Task. Vite+ wraps runtime management, package management, and frontend tooling in a single global CLI called `vp`. Vite+ is distinct from Vite
dograh-hq/dograh/AGENTS.md
agents with telephony and WebRTC support. ## Project Structure ``` dograh/ ├── api/ # Backend - FastAPI application ├── ui/ # Frontend - Next.js application ├── scripts/ # Helper scripts for local development ├── docs/ # Mintlify documentation ├── pipecat/ # Pipecat framework
dograh-hq/dograh/ui/AGENTS.md
Frontend Application Next.js 15 frontend for the Dograh voice AI platform. ## Project Structure ``` ui/ ├── src/ │ ├── app/ # Next.js App Router pages │ ├── components/ # React components │ ├── lib/ # Utilities and configurations │ ├── client/ # Auto-generated
go-vikunja/vikunja/AGENTS.md
Instructions Vikunja: self-hosted to-do app. Go API in `pkg/`, Vue 3 + TypeScript frontend in `frontend/` (pnpm). `veans/` is a separate Go module with its own `AGENTS.md`. ## Commands
go-vikunja/vikunja/veans/AGENTS.md
zero fails validation. Use `client.FarFuture` (year 9999) when you mean "no expiry" — the frontend does the same. - **Task descriptions and comments are HTML, not markdown.** The Vikunja web UI uses
No other project uses it yet
cloudflare/agents/agent-think/AGENTS.md
issue. - **Reproductions a human can see.** Every repro deploys with a minimal Vite frontend — click the `workers.dev` link, press _Trigger bug_, watch expected-vs-actual in the page (deployed
Steady3 mentions in 2 weeks
No other project uses it yet
An open format for instructions to coding agents, read by Codex, Cursor and others. Think of it as a README written for agents.
At the repository root, with more specific files in subdirectories. Agents read the one closest to the file they're editing.
Setup and test commands, code style, and the rules a new contributor would need to know.
Claude Code reads CLAUDE.md. A one-line CLAUDE.md that points at AGENTS.md covers both.