How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.
LumiwealthAGENTS.mdLumiwealth/lumibot/AGENTS.md
YYYY-MM-DD ### Added - New feature ### Changed - Modified behavior ### Fixed - Bug fix ### Deprecated / Removed / Security - As applicable ``` ### Pre-Deployment Checklist - [ ] Reviewed git commits since last version (`git log ..HEAD`) - [ ] Changelog
eugene1gAGENTS.mdeugene1g/agent-safehouse/AGENTS.md
every suite. - If adding policy behavior, add/update Bats coverage for the public contract or security boundary that changed. - Keep `.sb` header comments and any `#safehouse-test-id:*#` markers used
agentset-aiAGENTS.mdagentset-ai/agentset/.agents/skills/vercel-react-best-practices/AGENTS.md
invoked directly. Next.js documentation explicitly states: "Treat Server Actions with the same security considerations as public-facing API endpoints, and verify if the user is allowed to perform a mutation
rynfarAGENTS.mdrynfar/meridian/docs/agents.md
# Agent Setup This guide covers Claude client setup. For the Antigravity preview, use the
MicrosoftAGENTS.mdmicrosoft/intelligent-terminal/AGENTS.md
usage. ## Focused design references - Multi-window helper/master lifecycle: `doc/specs/Multi-window-agent-pane.md` - Session tracking: `doc/specs/hybrid-agent-session-tracking.md` - Security boundaries: `doc/security-model.md` - Installer: `doc/building-installer.md` - WTA customization: `tools/wta/CUSTOMIZATION.md
MicrosoftAGENTS.mdmicrosoft/intelligent-terminal/tools/wta/AGENTS.md
server: `src/cascadia/WindowsTerminal/TerminalProtocolComServer.cpp` Do not treat `WT_COM_CLSID`, `pipe-id`, or `set-env` as security boundaries. The human-facing WTA CLI intentionally does not expose direct keystroke injection; internal `ShellManager
doobidooAGENTS.mddoobidoo/mcp-memory-service/.claude/directives/agents.md
this repo. ## code-quality-guard **Purpose**: Fast automated analysis for complexity scoring, security scanning, and refactoring suggestions. **Capabilities:** - Complexity check (Gemini CLI or Groq API) - Security scan (SQL injection
WorldFlowAIAGENTS.mdWorldFlowAI/everything-claude-code/rules/agents.md
Test-driven development | New features, bug fixes | | code-reviewer | Code review | After writing code | | security-reviewer | Security analysis | Before commits | | build-error-resolver | Fix build errors | When build fails
xu-xiangAGENTS.mdxu-xiang/everything-claude-code-zh/AGENTS.md
智能体优先（Agent-First）** — 将领域任务委托给专业智能体处理 2. **测试驱动（Test-Driven）** — 先写测试后实现，要求 80% 以上的覆盖率 3. **安全第一（Security-First）** — 绝不在安全性上妥协；验证所有输入 4. **不可变性（Immutability）** — 始终创建新对象，绝不修改现有对象 5. **先规划后执行（Plan Before Execute）** — 在编写代码前先规划复杂功能的实现 ## 可用智能体
xu-xiangAGENTS.mdxu-xiang/everything-claude-code-zh/.codex/AGENTS.md
SKILL.md` — 详细说明与工作流（Workflow） - `agents/openai.yaml` — Codex 接口元数据 可用技能： - tdd-workflow — 测试驱动开发（TDD），包含 80% 以上的覆盖率 - security-review — 全面的安全自查清单 - coding-standards — 通用编码规范 - frontend-patterns — React/Next.js 模式 - frontend-slides — 视口安全（Viewport-safe
xu-xiangAGENTS.mdxu-xiang/everything-claude-code-zh/rules/agents.md
系统设计 | 架构决策 | | tdd-guide | 测试驱动开发 (TDD) | 新特性、Bug 修复 | | code-reviewer | 代码审查 | 代码编写/修改后 | | security-reviewer | 安全分析 | 提交代码前 | | build-error-resolver | 修复构建错误 | 构建失败时 | | e2e-runner | 端到端 (E2E) 测试 | 关键用户流程 | | refactor
MicrosoftAGENTS.mdmicrosoft/windows-drivers-rs/AGENTS.md
check --diff # Cargo.toml dependency sorting cargo sort -w -g -n --check # Spelling typos # Security audit cargo audit --deny warnings # Unused dependencies cargo machete --skip-target-dir # Docs build cargo
CoderGamesterAGENTS.mdCoderGamester/mcp-unity/AGENTS.md
token**: `Library/McpUnity/bridge-token` (256-bit hex secret; never commit or copy into project settings). - **Handshake security**: clients authenticate as `mcp-unity: ` with HTTP Basic and must omit `Origin`; every supplied Origin
NVIDIAAGENTS.mdNVIDIA/dcgm-exporter/deployment/AGENTS.md
features they enable, such as pod labels, pod UID, virtual GPU mapping, or DRA. - Security context changes are user-visible. Preserve the documented SYS_ADMIN/profiling tradeoff unless deliberately changing
MicrosoftAGENTS.mdmicrosoft/msphpsql/AGENTS.md
Windows SDK build: - **Visual Studio**: VS2019 for PHP 8.3, VS2022 for PHP 8.4+ - **Security flags**: `/NXCOMPAT`, `/DYNAMICBASE`, `/guard:cf`, `/ZH:SHA_256`, `/W4`, `/WX`, `/CETCOMPAT`, `/Qspectre` (for VCVERS >= 1913) - **Scripted
CloudflareAGENTS.mdcloudflare/workers-oauth-provider/AGENTS.md
this library provides both protected-resource middleware and the authorization server functionality needed to secure them. This library was largely written with Claude AI assistance, with all code thoroughly reviewed
Microsoft AzureAGENTS.mdAzure/azure-sdk-for-go/AGENTS.md
from specifications requires specific tools and workflows (see [code generation docs](https://github.com/Azure/azure-sdk-for-go/blob/main/documentation/development/generate.md)) #### Security and Compliance - **Security issues**: Must be reported privately to <secure@microsoft.com>, not in public issues - **License
u14appAGENTS.mdu14app/neo-chat/AGENTS.md
grouped under `src/components/`, with shared React hooks in `src/hooks/`. Shared domain logic, provider adapters, security gates, and API helpers live in `src/lib/`; client service wrappers are in `src/services/`; persisted Zustand
SentryAGENTS.mdgetsentry/sentry-react-native/AGENTS.md
loop before hypothesizing. | Warden's automated PR review and `agents.toml` also pull **remote** specialists (`security-review`, `gha-security-review`, `span-convention-review`, and more) — invoke those for depth beyond
DicklesworthstoneAGENTS.mdDicklesworthstone/pi_agent_rust/AGENTS.md
learn`) when `memory.backend` is `local` - Default-enabled: `jobs`, `hub` - `--tools` opt-in: `eval`, `github`, `security_scan` - Settings-gated: `browser`, `computer`, `inspect_image`, `generate_image`, `tts`, `read_media` - `subagent` - Native isolated
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.
