如何将 Apify Actors MCP Gateway 远程 MCP 连接到 OpenAI Codex CLI

约 1 分钟即可把 Apify 的远程 MCP 服务端接入 OpenAI Codex CLI。如需排查 OAuth / 401 问题,可先用 OAuth 2.1 调试器体检端点。

🔌 Gateway✓ VerifiedStreamable HTTPOAuth 2.1🛡️ A+~478 tok15 tools
Streamable HTTPhttps://mcp.apify.com

Add to ~/.codex/config.toml — Codex reads remote HTTP servers from [mcp_servers.*] tables:

[mcp_servers.apify]
url = "https://mcp.apify.com"

接入步骤

  1. 编辑 TOML 配置 — 打开 ~/.codex/config.toml,加入 [mcp_servers.apify] 与 url = "https://mcp.apify.com"。
  2. 配置凭据 — 首次调用时客户端会打开 Apify 的 OAuth 授权页面——点击允许后工具即可用。
  3. 验证 — 在 在线 Inspector 中粘贴 https://mcp.apify.com 并运行 initialize + tools/list 验证连通性。

🧰 工具预览(15)

ToolSafetyDescriptionParameters
search-actors
~30 tok
🟢 ReadSearch Apify Store for Actors.—
fetch-actor-details
~34 tok
🟢 ReadGet details and input schema of an Actor.—
call-actor
~30 tok
🟠 WriteRun an Actor and get its output.—
get-actor-output
~31 tok
🟢 ReadRetrieve output of an Actor run.—
add-actor
~31 tok
🟠 WriteAdd an Actor to the available toolset.—
remove-actor
~31 tok
🟠 WriteRemove an Actor from the toolset.—
get-actor-run
~30 tok
🟢 ReadGet details of an Actor run.—
get-actor-log
~29 tok
🟢 ReadGet logs of an Actor run.—
scrape-single-url
~32 tok
🟢 ReadScrape a single URL into markdown.—
fetch-apify-docs
~30 tok
🟢 ReadFetch Apify documentation.—
search-apify-docs
~30 tok
🟢 ReadSearch Apify documentation.—
get-dataset-items
~32 tok
🟢 ReadGet items from an Apify dataset.—
list-actor-runs
~29 tok
🟢 ReadList runs of an Actor.—
get-dataset
~28 tok
🟢 ReadGet details of a dataset.—
get-key-value-store
~31 tok
🟢 ReadGet a key-value store record.—

更多连接指南