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.
danielrosehillllms.txtdanielrosehill/Claude-Code-Projects-Index/docs/llms.txt
/danielrosehill/ai-tools-plugin): Early standalone Claude Code plugin with slash commands for AI tool management and MCP server configuration. - [Claude Code MCP Command Generator](https://github.com/danielrosehill/Claude-Code-MCP-Command-Generator): Generator for creating MCP server
io-ericllms.txtio-eric/coi/llms-full.txt
Pond/Drop) | | `coi self-upgrade` | Pull latest Coi source and rebuild the compiler | | `coi mcp` | MCP server for AI assistants: check, lookup, docs, search_packages, test | | `coi --out ` | Compile a single
io-ericllms.txtio-eric/coi/llms.txt
correct Coi syntax instead of JavaScript/TypeScript idioms. `coi mcp` is an MCP server with tools to check code, look up the exact APIs, read this guide by section, search packages
agnt-ggllms.txtagnt-gg/agnt/docs/llms-full.txt
createdAt": "2024-01-01T00:00:00Z" } ] } ``` ## MCP Routes Base path: `/api/mcp` ### Get All MCP Servers **GET** `/servers` - **Authentication**: Required - **Description**: Retrieve all MCP servers for the authenticated user - **Response
agnt-ggllms.txtagnt-gg/agnt/docs/llms.txt
evaluate, golden standards), Orchestrator (universal chat), Workflows (CRUD, start/stop, status), Executions, Tools, Custom Tools, MCP Servers (add, test, capabilities), NPM Search, Models (by provider, categories, cache), Streaming, Speech (Whisper transcription
Atharvsinh-codezllms.txtAtharvsinh-codez/ObsidianUI/public/llms-full.txt
readable HTTP API contract. - [Authentication](https://www.obsidianui.dev/markdown/authentication.md): Public read access requires no credentials. - [MCP server](https://www.obsidianui.dev/markdown/mcp.md): Local stdio resources/list and resources/read at obsidian://{name}. - [Registry catalogue
Atharvsinh-codezllms.txtAtharvsinh-codez/ObsidianUI/public/llms.txt
readable HTTP API contract. - [Authentication](https://www.obsidianui.dev/markdown/authentication.md): Public read access requires no credentials. - [MCP server](https://www.obsidianui.dev/markdown/mcp.md): Local stdio resources/list and resources/read at obsidian://{name}. - [Registry catalogue
Ikalus1988llms.txtIkalus1988/MisakaNet/docs/llms-full.txt
workflow-fix-dco-command-design-implementation](https://misakanet.org/lessons/dco-auto-fix-workflow-fix-dco-command-design-implementation/) - [Lesson: deepseek-tui-feishu-mcp-server-setup-permission-boundaries](https://misakanet.org/lessons/deepseek-tui-feishu-mcp-server-setup-permission-boundaries/) - [Lesson: deepseek-tui-write-file-sandbox-worktree-git-path
Ikalus1988llms.txtIkalus1988/MisakaNet/docs/llms.txt
database locked - Feishu document cleared - Windows Unicode crash - WSL permission denied - FANUC error codes - MCP server not discovered - Agent loops debugging ## How to connect ### Claude Code / Claude Desktop ```json { "mcpServers
Ikalus1988llms.txtIkalus1988/MisakaNet/llms.txt
repeat: - **MCP endpoint**: `https://misakanet.org/mcp` — Streamable HTTP, anonymous reads with no account. Reads carry a per-address burst window: a speed limit, not a quota. - **Local stdio server**: from
mixpeekPipes a download into a shellllms.txtmixpeek/amux/site/llms.txt
terminal TUI, git worktrees, lightweight, MIT); Nimbalyst (native macOS app, GPU rendering, no external server); Opcode (web-based, hosted, no local install); Claude Code Agent Teams (experimental native feature, requires
EvolutionAPIPipes a download into a shellllms.txtEvolutionAPI/evo-nexus/docs/llms-full.txt
with iteration-based circuit breaker ### Tier 2 — Setup & infrastructure (5) - `dev-mcp-setup` — configure MCP servers - `dev-deepinit` — generate hierarchical AGENTS.md / CLAUDE.md for new projects - `dev-project-session-manager
ratel-aiReads credentialsllms.txtratel-ai/ratel/llms.txt
user is running an MCP host (Claude Code, Cursor, ChatGPT) with multiple upstream MCP servers and wants one consolidated tool surface — point them at **ratel-local** ([`ratel-ai/ratel-mcp
oxbshwllms.txtoxbshw/LLM-Agents-Ecosystem-Handbook/llms.txt
safety, prompt evals + dataset + rubric ## Templates - [templates/](templates/): AGENTS, SOUL, MEMORY, USER, TOOLS, SKILL, MCP_SERVER, SYSTEM_PROMPT, AGENT_PROMPT, DESIGN_DOC, ADR, EVAL_PLAN, GUARDRAILS, HUMAN_APPROVAL_POLICY, CODING
bnb-chainReads credentialsllms.txtbnb-chain/bnb-chain.github.io/docs/llms.txt
privacy stack — Intelligent Privacy Pool, ZKredit, and Zama FHE. - [Overview](https://docs.bnbchain.org/developer-kit/mcp/): bnbchain-mcp server for on-chain tools and IDE integration. - [Ask AI](https://docs.bnbchain.org/developer-kit/ask-ai/): Read-only
nossa-yllms.txtnossa-y/activity-frames/llms.txt
aframes steps --frame ID | --find "task query" # replay-view script aframes mcp # MCP stdio server ## MCP (6 tools) get_context, get_activity, get_steps, get_day_summary, get_patterns
heyderekjllms.txtheyderekj/dinky/site/llms.txt
public HTTP API**, **no OAuth/OIDC**, **no protected API resource**, and **no remote MCP server** on dinkyfiles.com. - Machine-readable discovery: `https://dinkyfiles.com/openapi.yaml` (empty paths for the *site*; see description
BarredEwellms.txtBarredEwe/Prefire/llms.txt
also lets AI coding agents see SwiftUI: `prefire render` and the `prefire mcp` MCP server build the app and return previews as PNG images, on the devices, color scheme
erikdarlingdatallms.txterikdarlingdata/PerformanceMonitor/llms.txt
Server collectors, real-time alerts, graphical execution plan viewer, and a built-in MCP server for AI-powered analysis (more than 100 tools in Lite, more than 200 in Darling
ankimcpPipes a download into a shellllms.txtankimcp/anki-mcp-server/llms-full.txt
give AI systems the context they need. How it works 1 Choose MCP servers Pick from pre-built servers for popular tools like GitHub, Google Drive, Slack and hundreds
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.
