How real projects brief Codex, Cursor and every other agent that reads AGENTS.md.
sirkirbyAGENTS.mdsirkirby/unifi-mcp/AGENTS.md
controller_update(fields)` - Models live in `unifi-core` so both the MCP tool layer and the API server can import them - **Anchor:** `packages/unifi-core/src/unifi_core/network/models/acl.py` 2. Refactor tool functions to derive
veedstudioAGENTS.mdveedstudio/open-edit/AGENTS.md
they fit the ask. Use any video or image generating service and any available MCP server when it helps, not only the local engine and recipe path. Be creative: match
OpenAIAGENTS.mdopenai/build-hours/22-chatgptapp-pingpong/AGENTS.md
Always use the OpenAI developer documentation MCP server if you need to work with the OpenAI API, ChatGPT Apps SDK, Codex, without me having to explicitly
jrswabAGENTS.mdjrswab/axe/AGENTS.md
write = true # When true, write_file without explicit artifact arg writes to artifact dir [[mcp_servers]] name = "server-name" transport = "stdio" # or "http", "https" command = "/path/to/server" args = ["--flag
cyberkaidaAGENTS.mdcyberkaida/reverse-engineering-assistant/AGENTS.md
ReVa (Reverse Engineering Assistant) is a Ghidra extension that provides a Model Context Protocol (MCP) server for AI‑assisted reverse engineering. It offers a collection of tool providers that expose
TheGreenCedarAGENTS.mdTheGreenCedar/codex-autoresearch/AGENTS.md
skill. Do not restore retired subskills, slash-command documentation, MCP launchers, or a default MCP server without an explicit product-direction change and corresponding migration work. - The dashboard
llama-farmAGENTS.mdllama-farm/llamafarm/AGENTS.md
affected`) instead of using the underlying tooling directly - You have access to the Nx MCP server and its tools, use them to help the user - When answering questions about
utensilsAGENTS.mdutensils/mcp-nixos/AGENTS.md
with code in this repository. ## Project Overview MCP-NixOS is a Model Context Protocol (MCP) server that provides accurate, real-time information about NixOS packages, configuration options, Home Manager
BradGrouxAGENTS.mdBradGroux/veritas-kanban/AGENTS.md
Commander.js CLI (mirrors API endpoints) ├── shared/ Shared TypeScript types and utilities ├── mcp/ MCP server ├── desktop/ Electron desktop wrapper ├── docs/ Operator and developer documentation ├── prompt-registry/ Prompt templates and optional review
magicyuan876AGENTS.mdmagicyuan876/mineru-tianshu/AGENTS.md
Nginx） | 80 | | API 服务器 | 8000（注意：`.env.example` 中 `API_PORT=8080`，调试连接失败时先确认实际值） | | Worker | 8001 | | MCP Server | 8002 | | vLLM MinerU（手动 profile） | 30024 | ## 运行时架构 ### 进程拓扑 `start_all.py` 以子进程方式启动各服务；Docker 中每个服务是独立容器，共用 `tianshu
H-mmerAGENTS.mdH-mmer/pentest-agents/AGENTS.md
framework as a distributable product: - Keep `.claude/`, `rules/`, `skills/`, `tools/scaffold.py`, `tools/installer/`, provider bundles, MCP servers, tests, and docs aligned. - Treat `providers/` as generated output. Fix source or translators, then
indranilbanerjeeAGENTS.mdindranilbanerjee/digital-marketing-pro/AGENTS.md
Klaviyo, Amplitude, Canva, Figma, Asana, Webflow). User must configure env vars or set up MCP servers before they connect. 5. **Compliance is mandatory.** Skills enforce GDPR / CCPA
risa-labs-incAGENTS.mdrisa-labs-inc/BossConsole/AGENTS.md
requires equivalent isolation there. ### Running commands in a visible terminal pane When a terminal MCP server is available, prefer it over the plain `Bash` tool for commands worth showing
christopherkaraniAGENTS.mdchristopherkarani/Wax/AGENTS.md
only one of `wax-cli` / `wax-mcp` is built with embedder traits. Build both with `MiniLMEmbeddings` (and `MCPServer` for the server). ## Risk Areas - Dual Hermes trees drift unless `hermes_plugin_tests.sh
0xranxAGENTS.md0xranx/OpenContext/AGENTS.md
entry point (`oc`). - `src/core/` holds the Node.js core logic; `src/mcp/` hosts the MCP server. - `src/ui/` contains the Vite + React UI, with tests in `src/ui/tests/`. - `src-tauri/` and `crates
Archive228AGENTS.mdArchive228/loopkit/AGENTS.md
working tree. - `IMPLEMENTATION_PLAN.md` updated: what was done, what is next, known open issues. - Dev server killed (`./stop.sh` if the project has one). - A feature is only "done" after
controlplaneio-fluxcdAGENTS.mdcontrolplaneio-fluxcd/flux-operator/AGENTS.md
interacting with the Flux Operator. It is built using the Cobra library. * **Flux Operator MCP Server**: A server that connects AI assistants to Kubernetes clusters running the operator
dbt-labsAGENTS.mddbt-labs/dbt-agent-skills/evals/AGENTS.md
updates: **In `models.py`:** - `Grade` - grading result (success, score, tool_usage, notes, etc.) - `SkillSet` - skills, mcp_servers, allowed_tools - `Scenario` - name, path, prompt, skill_sets, description **In `runner.py`:** - `RunResult` - scenario results
GarethManningAGENTS.mdGarethManning/education-agent-skills/AGENTS.md
docs ## Test command ```bash npm test ``` ## After adding or editing any skill The MCP server serves a pre-built snapshot — changes to SKILL.md files will NOT appear on the live
msrbuildsAGENTS.mdmsrbuilds/elementor-mcp/AGENTS.md
lived MCP processes hold stale code.** After changing plugin PHP, the WP-CLI stdio MCP server the user is connected to keeps the old code until they reconnect
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.
