p5.js
processing/p5.js/AGENTS.md
analyze the codebase, DO NOT set up build environments, DO NOT run scripts, or MCP, or skills, DO NOT look things up online, DO NOT make code changes, UNTIL
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.
processing/p5.js/AGENTS.md
analyze the codebase, DO NOT set up build environments, DO NOT run scripts, or MCP, or skills, DO NOT look things up online, DO NOT make code changes, UNTIL
coleam00/Archon/AGENTS.md
parallel context switch. - Workflow AI nodes start without ambient skills, plugins, or MCP servers. A node receives only the skills, plugins, and MCP servers it explicitly names. Enforce that boundary
czlonkowski/n8n-mcp/AGENTS.md
configuration. ## Project Structure & Module Organization Core TypeScript lives in `src/`. Important subsystems include: - `src/mcp/` — MCP server, request handlers, tool definitions and documentation, and bundled skills - `src/database/` — SQLite adapters, repositories, FTS5
Steady9 mentions in 2 weeks
No other project uses it yet
claude-code-best/claude-code/AGENTS.md
DUMP_SYSTEM_PROMPT) - `--claude-in-chrome-mcp` / `--chrome-native-host` - `--computer-use-mcp` — 独立 MCP server 模式 - `--daemon-worker= ` — feature-gated (DAEMON) - `remote-control` / `rc` / `remote` / `sync` / `bridge` — feature-gated
1Panel-dev/MaxKB/AGENTS.md
JSON-RPC endpoint (`initialize`, `tools/list`, `tools/call`) routed through `chat.mcp.tools.MCPToolHandler`. MaxKB also *consumes* external MCP servers via `mcp_node` in the workflow and `langchain-mcp-adapters`. ### Frontend (`ui/`) Two SPAs built
PrimeIntellect-ai/prime-agent/AGENTS.md
prompts TOML rule map) followed by one dynamic tail (packages, project context, skills inventory, MCP servers, environment, session role); the harness digest stays a separate `[harness-digest]` user message. `prime
handsontable/handsontable/AGENTS.md
Angular, and Vue 3 wrappers. It runs entirely in the browser — frontend-only, no server-side logic. It cannot access the internet unless explicitly configured (air-gapped environment support). There
QwikDev/qwik/AGENTS.md
repo. Use `~/.config/ruler/` for personal preferences, local workflow shortcuts, API keys, and personal MCP servers: ```bash ruler init --global ``` If it helps everyone working in this repo
QwikDev/qwik/.ruler/AGENTS.md
generate them (once per worktree): > > ```bash > pnpm dlx @intellectronica/ruler@0.3.42 apply --no-gitignore --no-mcp > ``` > Canonical source for repo-wide AI coding agent rules. For contributor setup, see > [CONTRIBUTING.md](./CONTRIBUTING.md
github/docs/content/copilot/responsible-use/agents.md
Copilot. The SDK communicates with Copilot CLI over JSON-RPC and supports custom agents, MCP server integrations, lifecycle hooks, and session management. * **The {% data variables.copilot.github_copilot_app %}**: A desktop application
pydantic/pydantic-ai/AGENTS.md
OTel instrumentation of Pydantic AI and `httpx` - If you have access to the Logfire MCP server, you can use it to inspect agent runs, tool calls, and model requests ## Test
pydantic/pydantic-ai/src/pydantic_clai2/AGENTS.md
owns the host API, loading, and credential picker helpers. `builtin_plugins/` and `mcp/` retain their plugin and server boundaries. Move a capability to core or harness if it does
pydantic/pydantic-ai/tests/AGENTS.md
without proper assertions or that verify opposite behavior create false positives - Test MCP against real `tests.mcp_server` instance, not mocks — extend test server with helper tools to expose runtime context
gofr-dev/gofr/docs/AGENTS.md
gofr.dev/docs/advanced-guide/injecting-databases-drivers> - Calling LLMs (`app.AddLLM`, `c.LLM()`, env config, multiple models, embeddings): <https://gofr.dev/docs/advanced-guide/llm> - MCP server and agent loops (`app.EnableMCP`, `c.LLM().Tools()`): <https://gofr.dev/docs/advanced-guide/mcp> - Streaming responses (SSE / NDJSON, `response.Stream
storytold/pdfcraft/AGENTS.md
Never enable it by default, widen the bind address, or log the token. - **The MCP server is opt-in.** Never make the app, the CLI or an installer start
JetBrains/intellij-community/AGENTS.md
Tools Search the repository with `./tools/accio.cmd`, not with the `search_*` tools of an IDE MCP server. It answers from the working tree of the current checkout, a lease included
kortix-ai/CompanyOS/AGENTS.md
with `runtimeSupports`. pi lists `session.subagents`, `session.compact`, `session.commands` and `session.steer`. pi does not serve rewind, MCP servers, the todo list, shell turns, part edits or `session.attach`. The harness rules
refactoringhq/tolaria/AGENTS.md
obtain repository-owner approval before pushing. **If CodeScene gate blocks your push:** use `mcp__codescene__code_health_score` to find the worst file, refactor it, commit, push again
teng-lin/notebooklm-py/AGENTS.md
Internal:** `_*.py` and `_*/` (`_app/`, `_auth/`, `_runtime/`, `_sources.py`, `_artifacts.py`). - **Adapters & Wire:** `cli/` (Click), `mcp/` (FastMCP), `server/` (FastAPI), `rpc/` (batchexecute RPC facade), `_android/` (gRPC session & codecs). - **Tests & Data:** `tests/unit/`, `tests/integration/` (VCR), `tests/server
Steady5 mentions in 2 weeks
No other project uses it yet
rocketride-org/rocketride-server/AGENTS.md
catalog) | | Test a node | `docs/development/nodes/testing.md` | | Client (SDK) docs | `docs/development/clients/readme-schema.md`; sources in `docs/public/ /` | | MCP surface | `docs/public/mcp/` | | Engine internals | `docs/development/engine/index.md` | | WebSocket (5565) protocol | `docs/public/product/connect/websocket/` | | Crash reporting / minidumps | `docs/development/engine/crash-reporting.md` | | Shell apps | `docs/public/product/guides/apps/index.md
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.