zedis
vicanso/zedis/AGENTS.md
RBAC, MCP, `/v1/watch`, SQL Stores, Page Data)**: `CLAUDE.md` line 35, lines 60–84 - **Nothing in the GUI Crate Speaks Redis (ADR 10)**: `CLAUDE.md` lines 85–101 - **Architecture, Server Feature Probe
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.
vicanso/zedis/AGENTS.md
RBAC, MCP, `/v1/watch`, SQL Stores, Page Data)**: `CLAUDE.md` line 35, lines 60–84 - **Nothing in the GUI Crate Speaks Redis (ADR 10)**: `CLAUDE.md` lines 85–101 - **Architecture, Server Feature Probe
wesm/msgvault/AGENTS.md
hardware. The daemon owns archive access; the CLI, Web UI, TUI, HTTP API, and MCP server provide different ways to use the same archive. - [Architecture](docs/architecture/overview.md): responsibilities and data flow
alpic-ai/skybridge/AGENTS.md
MCP Apps — interactive React views that render inside AI conversations. The core loop: an MCP server exposes tools. When the host (ChatGPT, Claude, VSCode…) calls a tool, the server returns
Steady2 mentions in 2 weeks
No other project uses it yet
microsoft/intelligent-terminal/AGENTS.md
recovered. - Terminal mutation requested by an agent goes through the confirmation-gated session MCP action path. Agent-owned shell tools are a separate execution path. ## Key files | Area | Path | | --- | --- | | Terminal
Steady2 mentions in 2 weeks
No other project uses it yet
microsoft/intelligent-terminal/tools/wta/AGENTS.md
processes or invent process controls for them. ### Session MCP Master owns one loopback Streamable HTTP listener and publishes an independent server name and bearer capability for each eligible ACP session
Steady2 mentions in 2 weeks
No other project uses it yet
isair/jarvis/AGENTS.md
runs the same router; never removes stop/self; capped per reply | | `src/jarvis/tools/external/mcp_runtime.spec.md` | Persistent MCP runtime: per-server long-lived stdio session, queue-based dispatch, retry on transient session loss | One worker
apache/tinkerpop/AGENTS.md
Cursor:gpt-5`, or `Assisted-by: Claude Code:claude-opus-4-7 [tinkerpop-mcp]`. This aligns with the ASF's recommendation on AI provenance tracking. `Assisted-by:` is the *only
wuyoscar/AISafetyHot-Hub/AGENTS.md
connection guide; `docs/mcp-examples.md`: usage examples. - `skills/*/SKILL.md`: the agent skills (all backed by the MCP server); `.claude-plugin/` and `.mcp.json`: the Claude Code plugin and marketplace. - `assets/`: branding and support
zone-eu/wildduck/AGENTS.md
server and REST API backed by MongoDB and Redis. It runs as a standalone server and as a shared library imported through `@zone-eu/wildduck/lib/*` by other mail components
feigeCode/navop/AGENTS.md
# 全局 Agent 规则 本文件用于约束自动化代理在本机工作区中的默认工作方式,强调最短路径、风险分级、TDD 思想与可验证交付。 ## 指令优先级 - 指令优先级从高到低: 1. 当前会话中用户的明确要求 2. 仓库自身的规则
JMBeresford/retrom/.agents/AGENTS.md
committed). Each is a directory with SKILL.md. - skills-local/ — Private skills (gitignored). - mcp/servers.json — Canonical MCP server config. - scripts/ — Sync and validation scripts (link-skills.sh, sync-mcp.sh). Adding a skill 1. Create `.agents/skills
WuKongOpenSource/AI_CRM/AGENTS.md
MySQL 数据同步到 AI CRM PostgreSQL,默认端口为 `10456`;它不使用 backend 的 Services、认证或租户拦截器,也不会在启动时自动同步。 - `mcp_server/` 是独立的 stdio MCP 服务器,会调用 backend HTTP API;它有意不调用也不暴露 `sync_data`。 - `CLAUDE.md` 是面向架构的综合参考(仓库结构、命令
trycompai/comp/AGENTS.md
customer-facing endpoint in `apps/api/src/` flows into three systems: the OpenAPI spec (`packages/docs/openapi.json`), the MCP server published as `@trycompai/mcp-server` on npm, and the runtime `ValidationPipe`. The full contract
zerx-lab/warp/AGENTS.md
枚举与 `DOGFOOD/PREVIEW/RELEASE_FLAGS` | | `warp_util` | 跨多个 crate 复用的通用工具函数 | | `warp_logging` | 日志配置统一入口 | | `simple_logger` | 给 `remote_server` 等 stderr-only 进程用的简易异步文件日志 | | `warp_web_event_bus` | Web 端事件总线(给嵌入的 web view) | | `field_mask
ai4s-research/open-science/AGENTS.md
sidecar. - Keep the frontend, desktop shell, and agent runtime decoupled. - Skills, MCP servers, and model providers must stay pluggable. - Keep the artifact schema and workflow templates stable and versioned. - Every
GoogleCloudPlatform/scion/web/AGENTS.md
using the Chrome DevTools MCP tools (`take_screenshot`, `take_snapshot`, `click`, `fill`, etc.) to validate the web UI, follow these guidelines. ### Starting the Go server 1. **Settings must exist first
No other project uses it yet
coder/cmux/docs/AGENTS.md
Agent instructions for AI assistants working on the Xum codebase
coder/mux/docs/AGENTS.md
Agent instructions for AI assistants working on the Xum codebase
scaffold-eth/scaffold-eth-2/AGENTS.md
library (Wagmi, Viem, RainbowKit, DaisyUI, Hardhat, Next.js, etc.). Context7 is configured as an MCP server and provides access to indexed documentation with code examples. ## Skills & Agents Index IMPORTANT: Prefer retrieval
ezEngine/ezEngine/AGENTS.md
audio, ...) - `Code/Editor/` - editor application and framework, `ezEditorProcessor` - `Code/EditorPlugins/` - editor plugins (asset types, scene editing, MCP server, ...) - `Code/Tools/` - standalone tools - `Code/UnitTests/` - test executables (FoundationTest, CoreTest, GameEngineTest, RendererTest, EditorTest, ...) - `Code/ThirdParty/` - third-party
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.