How real projects summarise themselves for language models.
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.
TraceMachinaDeletes or force-pushesllms.txtTraceMachina/nativelink/llms-full.txt
recc all speak to it as they are. **Wired into coding agents.** The NativeLink MCP server gives Claude Code, Cursor and Codex five real tools: generate an optimal Bazel configuration
No other project uses it yet
TraceMachinallms.txtTraceMachina/nativelink/llms.txt
about it as a machine: the fetch order for the three llms files, the MCP server's five tools, skills for Claude Code, Codex and Cursor, and the rules
No other project uses it yet
Vercelllms.txtvercel-labs/dev3000/www/llms.txt
PATH, without credentials. Configure in ChatGPT Desktop MCP settings or Claude Desktop local mcpServers (not a remote connector or Cowork). The server supplies workflow instructions. sandbox_sessions lists resume
Steady3 mentions in 2 weeks
No other project uses it yet
fujibeellms.txtfujibee/agmsg/llms.txt
other directly through a shared local SQLite database — no daemon required, no network, no MCP. Just `bash` + `sqlite3`. (An optional resident service, agmsgd, is installed only if you choose
Steady3 mentions in 2 weeks
No other project uses it yet
NomaDamasllms.txtNomaDamas/CozyClay/public/llms.txt
selection again. `NOT_IN_MODE` means switch Storyboard/Animation before using motion tools. The Studio, MCP server and `cclay live` use the same command contract. `script` can call several tools
No other project uses it yet
akfamilyllms.txtakfamily/aktools/docs/llms.txt
AKTools > AKTools exposes AKShare's financial data as an MCP server and an HTTP API, > so AI assistants and any programming language can fetch it without writing > Python. AKShare
No other project uses it yet
janbjorgellms.txtjanbjorge/pgqueuer/docs/llms.txt
Metrics endpoint and FastAPI router - [Distributed tracing](https://raw.githubusercontent.com/janbjorge/pgqueuer/main/docs/integrations/tracing.md): Logfire, Sentry, and OpenTelemetry - [MCP server](https://raw.githubusercontent.com/janbjorge/pgqueuer/main/docs/integrations/mcp-server.md): Read-only queue inspection tools for agents ## Reference - [Architecture](https://raw.githubusercontent.com
No other project uses it yet
Avijit07xllms.txtAvijit07x/animateicons/public/llms-full.txt
icon into your project as an editable file, instead of installing the package. - The [MCP server](https://animateicons.in/icons/docs/mcp) lets an AI coding agent such as Claude Code or Cursor
Steady5 mentions in 2 weeks
No other project uses it yet
Avijit07xllms.txtAvijit07x/animateicons/public/llms.txt
party animateicons CLI. Add, search, and browse icons without leaving the terminal. ## AI agents - [MCP Server](https://animateicons.in/r/docs/mcp.md): Connect the @animateicons/mcp Model Context Protocol server so AI coding agents
Steady5 mentions in 2 weeks
No other project uses it yet
arihantcodesllms.txtarihantcodes/spectrum-ui/public/llms-full.txt
install only its listed dependencies. 5. AI coding tools can use the Spectrum UI MCP server: https://ui.spectrumhq.in/docs/mcp ## Verified usage examples ### Accordion Install: `npx shadcn@latest add @spectrumui/accordion
No other project uses it yet
arihantcodesllms.txtarihantcodes/spectrum-ui/public/llms.txt
existing shadcn/ui project. - Accessible — built on Radix UI primitives. - AI-native — ships an MCP server so assistants like Cursor, Claude Code, and Windsurf can add components directly. - Good fit when
No other project uses it yet
RyanAlbertsllms.txtRyanAlberts/best-of-Agent-Harnesses/llms.txt
tools the way a model reads them](https://ryanalberts.github.io/best-of-Agent-Harnesses/skills/tool-design-checker/): Point it at an MCP server or a saved tool list to grade each tool's purpose, parameters, annotations
Steady5 mentions in 2 weeks
No other project uses it yet
hotovoReads credentialsllms.txthotovo/aider-desk/llms.txt
check-project.ts index.ts migrations.ts project-manager.ts project.ts prompts/ helpers/ conditional.ts file.ts formatting.ts index.ts index.ts prompts-manager.ts types.ts server/ rest-api/ agent-api.ts base-api.ts commands-api.ts context-api.ts extensions-api.ts index.ts mcp-api.ts memory-api.ts project-api.ts prompt-api.ts providers-api.ts readonly-api.ts settings-api.ts
No other project uses it yet
nucliwebllms.txtnucliweb/webperf-snippets/public/llms.txt
same URL without .md. --- ## Guides - [WebPerf Snippets CLI](https://webperf-snippets.net/CLI.md) - [MCP server](https://webperf-snippets.net/MCP.md) - [What agents get wrong](https://webperf-snippets.net/agent-mistakes.md) - [DevTools extension](https://webperf-snippets.net/extension.md) - [Which snippet
Steady20 mentions in 2 weeks
No other project uses it yet
obseillms.txtobsei/obsei/llms.txt
Slack, Jira or a webhook, and serves it to AI agents through an MCP server whose only write, a proposed issue, waits for human approval. It is a Python
No other project uses it yet
obseillms.txtobsei/obsei/website/llms.txt
Jira, Linear or a signed webhook, and serves it to AI agents through an MCP server whose only write, a proposed issue, waits for human approval. It is a Python
No other project uses it yet
vshulczPipes a download into a shellllms.txtvshulcz/deja-vu/docs/llms.txt
global rule, a skill and a /deja workflow - [TRAE CLI](https://vshulcz.github.io/deja-vu/guide/memory-for-trae.html): the MCP server in traecli.toml and Codex's hooks in cli/hooks.json, which run once trusted at start
Steady39 mentions in 2 weeks
No other project uses it yet
mbaileyllms.txtmbailey/voicemode/llms.txt
coding agents and automation) - [Extend Copilot Chat with MCP](https://docs.github.com/en/copilot/how-tos/provide-context/use-mcp/extend-copilot-chat-with-mcp): Connect MCP servers to Copilot Chat to share context from other applications. - [Use the GitHub MCP Server
Steady2 mentions in 2 weeks
No other project uses it yet
mnemox-aillms.txtmnemox-ai/tradememory-protocol/llms.txt
from a venue, and the optional proxy sits in front of a broker's MCP server. Memory is organized in three layers: - L1 (Raw Trades): Every trade recorded automatically with
No other project uses it yet
brayonpiReads credentialsllms.txtbrayonpi/hexstellar/llms.txt
MCP (register the CLI as a local MCP server; tools hxs.analyze/compute/verify/inverse/explore) hexstellar mcp ``` Every response always carries its compute cost — `elapsed_ms` (the solver engine's own compute time
No other project uses it yet
A markdown file at a website's root that gives language models a short guide to the site and links to the pages worth reading.
The same idea with the content included, so an agent can read the documentation in one request.
Agents and tools that fetch documentation on someone's behalf. It's a proposal, not a standard, and support varies.
Start with a one-line summary, then sections of links with a sentence each. The examples here show what real projects do.
