# DotsAgent

> LLM API prices, an agent cost calculator, MCP configs for 13 clients, AGENTS.md and llms.txt generators, and agent security notes. Free, in 18 languages.

Facts checked 2026-10-01. Every page has a Markdown version at /md/<path>/index.md and is also served as Markdown when requested with Accept: text/markdown. Pages exist in 18 languages under /<lang>/ (English at the root).

## Build

- [Models and API prices](https://dotsagent.io/md/models/index.md): Input, cached and output prices per million tokens for 28 models from OpenAI, Anthropic, Google, xAI, DeepSeek, Qwen and more, with context sizes and agent features.
- [Agent cost calculator](https://dotsagent.io/md/cost/index.md): Estimate what one agent task and a month of tasks cost on 28 LLM APIs. Set prompt size, context growth, steps, output and cache hit rate; long-prompt rates included.
- [Frameworks and SDKs](https://dotsagent.io/md/frameworks/index.md): LangGraph, CrewAI, OpenAI Agents SDK, Google ADK, Mastra and 11 more: languages, licences, versions, MCP, durable runs, approvals and evals, checked 1 Oct 2026.

## Connect (MCP)

- [MCP servers](https://dotsagent.io/md/mcp/index.md): What changed in MCP spec 2026-07-28, how auth and the official registry work, and setup details for 60 checked servers across 13 clients.
- [MCP config generator](https://dotsagent.io/md/mcp/config/index.md): Pick servers and a client to get its config file in the right shape: mcp.json, config.toml, YAML or settings.json, plus CLI commands. 13 clients, 60 servers.
- [MCP clients](https://dotsagent.io/md/mcp/clients/index.md): Where 13 MCP clients keep their servers, from Claude Code to Zed: config paths, top-level keys, formats, remote OAuth support and the quirks that break setups.

## Instruct

- [AGENTS.md generator](https://dotsagent.io/md/instructions/index.md): Fill in one form to get an AGENTS.md, the same path-scoped rule for Cursor, Copilot, Claude Code, Cline and Devin, and bridge files for Gemini CLI and Aider.
- [Instruction files](https://dotsagent.io/md/instructions/files/index.md): Where AGENTS.md, CLAUDE.md, GEMINI.md and the Cursor, Copilot, Devin and Cline rule files live, how they merge, and which of 9 coding tools read AGENTS.md.

## Run

- [Self-host OpenClaw](https://dotsagent.io/md/openclaw/index.md): Self-host OpenClaw 2026.9.7 on a laptop, VPS, Docker or Raspberry Pi, connect models and Telegram, and keep port 18789 private with the hardened baseline.

## Agent-ready websites

- [Agent-ready websites](https://dotsagent.io/md/agent-ready/index.md): Generate llms.txt, a robots.txt for AI bots and an RFC 9727 API catalog, then look up 40 AI crawlers and agents by token, purpose and published IP list.
- [llms.txt generator](https://dotsagent.io/md/agent-ready/llms-txt/index.md): Generate an llms.txt file in the llmstxt.org format, with an H1 name, a blockquote summary and H2 link sections. Follows the 10 August 2026 spec revision.
- [robots.txt for AI bots](https://dotsagent.io/md/agent-ready/robots-txt/index.md): Build a robots.txt that refuses AI training but keeps search: 35 bot tokens grouped by purpose, Content-Signal lines, private paths and a sitemap line.
- [API catalog generator](https://dotsagent.io/md/agent-ready/api-catalog/index.md): Build the RFC 9727 JSON linkset for /.well-known/api-catalog, listing each API with its OpenAPI description, docs and status page, plus an nginx block.
- [AI bot directory](https://dotsagent.io/md/agent-ready/bots/index.md): 40 AI crawlers and agents from OpenAI, Anthropic, Google, Apple, Meta, Amazon and others, with robots.txt token, purpose, compliance and IP list.

## Secure

- [Agent security](https://dotsagent.io/md/security/index.md): Agent security reference: both 2026 OWASP top 10 lists, 17 real incidents from MCP to OpenClaw, 15 defences and a checklist for teams shipping AI agents.
- [Incident log](https://dotsagent.io/md/security/incidents/index.md): Public AI agent security incidents since April 2025, from MCP tool poisoning to Claude Code WebFetch exfiltration, each with CVEs, root cause and fixes.
- [Security checklist](https://dotsagent.io/md/security/checklist/index.md): A 22-item security checklist for AI agents covering untrusted input, tool credentials, sandboxing, MCP tokens and CI secrets. Progress is saved in your browser.

## Track

- [Changelog](https://dotsagent.io/md/changelog/index.md): Dated log of what changed for agent builders: model launches and retirements, API and SDK releases, MCP spec updates and price moves, each with sources.

## Optional

- [Sitemap](https://dotsagent.io/sitemap.xml): sitemap index with one sitemap per language
- [Español](https://dotsagent.io/es)
- [Português](https://dotsagent.io/pt)
- [Français](https://dotsagent.io/fr)
- [Deutsch](https://dotsagent.io/de)
- [Italiano](https://dotsagent.io/it)
- [Nederlands](https://dotsagent.io/nl)
- [Polski](https://dotsagent.io/pl)
- [Türkçe](https://dotsagent.io/tr)
- [Русский](https://dotsagent.io/ru)
- [العربية](https://dotsagent.io/ar)
- [简体中文](https://dotsagent.io/zh)
- [日本語](https://dotsagent.io/ja)
- [한국어](https://dotsagent.io/ko)
- [Bahasa Indonesia](https://dotsagent.io/id)
- [Tiếng Việt](https://dotsagent.io/vi)
- [ไทย](https://dotsagent.io/th)
- [Filipino](https://dotsagent.io/fil)