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.
bibinprathapllms.txtbibinprathap/VeritasGraph/llms.txt
endpoints - GraphRAG pipeline components ## Canonical Paths - Repo root README: [README.md](./README.md) - Cookbook: [cookbook/](./cookbook/) - MCP server (zero-trust, air-gapped GraphRAG over JSON-RPC stdio): [veritasgraph_mcp/](./veritasgraph_mcp/) - Studio backend
SveltyCMSReads credentialsllms.txtSveltyCMS/SveltyCMS/static/llms.txt
gates. ## Machine-Readable Extras - [OpenAPI Spec](https://demo.sveltycms.com/api/openapi.json): Live generated schema. - [MCP Knowledge Server](https://mcp.sveltycms.com/mcp): Model Context Protocol endpoints for AI agent tool integration. - [GitHub Repository
rife2llms.txtrife2/bld/llms.txt
/rife2/bld/wiki/Custom-Commands): add a project command by writing a method and annotating it with @BuildCommand - [MCP server](https://github.com/rife2/bld/wiki/Standard-Commands#mcp): serve every build command as an MCP tool over stdio, register
n1m21nllms.txtn1m21n/Infinite/website/llms-full.txt
drawn on the picture or mixed into the sound. - Every result from the MCP server and from the command line with --json includes an `attribution` object. Its `credit_line
n1m21nllms.txtn1m21n/Infinite/website/llms.txt
drawn on the picture or mixed into the sound. - Every result from the MCP server and from the command line with --json includes an `attribution` object. Its `credit_line
vishalsachdevllms.txtvishalsachdev/canvas-mcp/docs/llms.txt
Canvas MCP > Canvas MCP is an open-source Model Context Protocol (MCP) server that bridges AI assistants with Canvas Learning Management System (LMS). It provides up to 128 tools
GabrieleLozuponellms.txtGabrieleLozupone/kokon/site/llms.txt
before/after compare, private mode, works offline after setup. - For AI agents: built-in MCP server (generate_image, edit_image, inpaint_image, outpaint_image, upscale_image, ...) and a local HTTP
HigangsshPipes a download into a shellllms.txtHigangssh/homebutler/llms.txt
network scan, Wake-on-LAN - Multi-server SSH management and remote deploy - Built-in MCP server for AI tools - JSON output for scripts and agents ## Primary workflow For monitoring
Hugging Facellms.txthuggingface/paperswithcode/frontend/public/llms.txt
papers outside the catalog. ## Links - API schema: https://paperswithcode.co/docs - MCP specification: https://modelcontextprotocol.io/docs/2026-07-28/learn/server-concepts - CLI and server source: https://github.com/huggingface/pwc-cli
SonarSourcePipes a download into a shellllms.txtSonarSource/sonarqube-cli/docs/llms.txt
integration for Claude Code. This will install secrets scanning hooks, configure Vortex analysis and MCP Server. Usage: sonar integrate claude [--non-interactive] [--global] Options: --non-interactive Non-interactive mode
Keesan12llms.txtKeesan12/martin-loop/llms-full.txt
compatible agent hosts use the MartinLoop workflow directly. Install: `npx -y @martinloop/mcp@latest` The MCP server uses local stdio transport for the public package. Through MCP, a host
Undertone0809llms.txtUndertone0809/rudder/docs/llms.txt
future Agent work. - [Plugins](https://docs.rudderhq.dev/concepts/plugins): Install Codex-compatible packages that combine Skills, MCP servers, and Apps. ### How-to guides - [Use an Issue from Start to Finish](https://docs.rudderhq.dev
vidispinellms.txtvidispine/hull/llms.txt
result - [Tools for AI agents](https://raw.githubusercontent.com/vidispine/hull/main/hull/files/doc/agent_tools.md): the Claude Code plugin, the MCP server and hull_validate.py - [Transformations](https://raw.githubusercontent.com/vidispine/hull/main/hull/files/doc/transformations.md): _HT*, _HT!, _HT?, _HT^, _HT&, _HT/ in detail - [Chart
AGIHuntPipes a download into a shellllms.txtAGIHunt/blurt/site/llms.txt
worktree per item. Review the changes, then merge. - Agents report through Blurt's MCP server: In progress, then To verify. MCP tools: search_items, get_item, get_transcript, update_item
noskillishllms.txtnoskillish/bankmcp/docs/llms.txt
BankMCP™ > BankMCP™ (also written Bank MCP or bankmcp) is an open-source MCP server for your own bank accounts. You run it yourself. It connects to your banks through Enable
Scottcjnllms.txtScottcjn/rustchain-bounties/llms.txt
onramp — 4-step flow from pick → claim → ship → get paid. - [Scottcjn/bounty-concierge](https://github.com/Scottcjn/bounty-concierge): MCP server exposing the board to your agent loop. ## Claim a bounty 1. Browse [open bounty
jfarcandllms.txtjfarcand/mirroir-mcp/llms.txt
mirroir-mcp > Give your AI eyes, hands, and a real iPhone. An MCP server that lets any AI agent see the screen, tap what it needs, and figure the rest
Continuum-AI-Corpllms.txtContinuum-AI-Corp/OrcaReplay/llms.txt
against your own traces. ## MCP - [server.json](https://github.com/Continuum-AI-Corp/OrcaReplay/blob/main/server.json): MCP server manifest. `orca mcp` serves the trace store over stdio as six tools, so an agent can list, read, diff
nirholasReads credentialsllms.txtnirholas/three.ws/llms-full.txt
brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native. # three.ws [**Website**](https://three.ws) · [**Docs**](https://three.ws/docs) · [**Changelog
nirholasllms.txtnirholas/three.ws/llms.txt
brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native. three.ws is an open-source repository by nirholas, part
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.
