mcp
Adios.dev
Documentation
Adios MCP
Use Adios workspaces, previews, builds, logs, and deployments from Codex, ChatGPT, and other MCP clients.
This repository packages the hosted Adios MCP service for public installation. It includes a Codex marketplace plugin, two reusable skills, connection examples, and a readable tool reference. The server runs at Adios, so users don't need to deploy anything from this repository.
Endpoints
| Purpose | URL |
|---|---|
| Streamable HTTP (preferred) | `https://api.adios.dev/v1/mcp` |
| SSE fallback | `https://api.adios.dev/v1/mcp/sse` |
| Public connection metadata | `https://api.adios.dev/v1/mcp/config` |
| Public tool discovery | `https://api.adios.dev/v1/mcp/tools` |
Tool calls use Adios OAuth and are scoped to the authenticated user and team. Connection metadata and tool schemas are public; an Adios account is required to act on workspaces or workloads.
Install in Codex
Register the GitHub marketplace:
codex plugin marketplace add adiosdotdev/mcpRestart the ChatGPT desktop app, open Plugins, choose the Adios marketplace, install Adios MCP, and complete OAuth when prompted. The plugin installs the remote MCP connection and both workflow skills.
To connect only the MCP server without the plugin:
codex mcp add adios --url https://api.adios.dev/v1/mcp
codex mcp login adiosThe equivalent `config.toml` is in `examples/codex-config.toml`.
Install only the skills
Ask Codex to install either GitHub path with `$skill-installer`:
Use $skill-installer to install https://github.com/adiosdotdev/mcp/tree/main/plugins/adios-mcp/skills/adios-build-deployUse $skill-installer to install https://github.com/adiosdotdev/mcp/tree/main/plugins/adios-mcp/skills/adios-operate-debugSkill-only installs do not configure the MCP server; add the `adios` MCP connection separately.
Use through npm
For MCP clients that start a local stdio command, use the public npm package:
{
"mcpServers": {
"adios": {
"command": "npx",
"args": ["-y", "@adiosdotdev/mcp"]
}
}
}The command starts a local bridge to the hosted Adios endpoint and opens OAuth when the client connects. Nothing from the Adios platform runs locally.
Other MCP clients
Create a remote or Streamable HTTP MCP connection named `adios` with this URL:
https://api.adios.dev/v1/mcpLet the client complete OAuth rather than adding a static token. If the client does not support Streamable HTTP, use the SSE endpoint shown above. Configuration field names vary by client.
Included workflows
- `adios-build-deploy`: create or import a workspace, update source, validate `adios.yaml`, run lint/build, start a preview, and deploy after approval.
- `adios-operate-debug`: inspect Git/runtime/command state, analyze logs, diagnose failures, and perform verified operational fixes.
See `docs/tool-reference.md` for the current tool groups and approval-sensitive operations. The live `/tools` endpoint remains the source of truth.
Safety
Deploys, workload changes, runtime stops, and file deletion must remain approval-sensitive. Never commit Adios OAuth tokens, app credentials, or workspace secrets to this repository or to project source.
Repository layout
.agents/plugins/marketplace.json Codex marketplace catalog
plugins/adios-mcp/.codex-plugin/ Plugin manifest
plugins/adios-mcp/.mcp.json Hosted MCP connection
plugins/adios-mcp/skills/ Reusable Codex workflows
bin/adios-mcp.js npm stdio bridge command
docs/ Tool and operating reference
examples/ Direct client configurationAdios platform documentation: www.adios.dev/mcp
License
MIT
Frequently asked questions
What is mcp?
mcp is Adios.dev
How do I install mcp?
Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.
Is mcp open source?
Yes — it is hosted on GitHub at https://github.com/adiosdotdev/mcp.
Related MCP tools
🔥 Official Firecrawl MCP Server - Adds powerful web scraping and search to Cursor, Claude and any other LLM clients.
Use any LLMs (Large Language Models) for Deep Research. Support SSE API and MCP server.
Enhanced MCP server for interactive user feedback and command execution in AI-assisted development, featuring dual interface support (Web UI and Desktop Application) with intelligent environment detection and cross-platform compatibility.
A powerful Zotero AI and MCP plugin with ChatGPT, Gemini 3.7, Claude Fable 5, Claude Opus 5, DeepSeek V4, Grok, OpenRouter, Kimi k3, GLM 5.3, SiliconFlow, GPT-oss, Gemma 4, Qwen 3.8
Connect your browser to AI models. Just use Dia on Chrome, Arc or Firefox.
文颜 MCP Server 可以让 AI 自动将 Markdown 文章排版后发布至微信公众号。
Run your own MCP server? See who uses it and what to fix.
Measure it with TrackMCP