{"id":8498,"date":"2026-05-25T06:45:53","date_gmt":"2026-05-24T22:45:53","guid":{"rendered":"https:\/\/infernews.com\/blog\/post-bfdf0ad9\/"},"modified":"2026-05-25T06:49:16","modified_gmt":"2026-05-24T22:49:16","slug":"post-bfdf0ad9","status":"publish","type":"post","link":"https:\/\/infernews.com\/blog\/post-bfdf0ad9\/","title":{"rendered":"AIQ \u70ba\u4ee3\u7406\u5de5\u5177\u52a0\u4e0a\u6df1\u5ea6\u7814\u7a76\u80fd\u529b"},"content":{"rendered":"\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/uploads\/2026\/05\/Agentic-AI-2a2d1b018f63.jpg\" alt=\"Og image\"\/><\/figure>\n\n\n\n<p>\u4e0d\u5c11\u4ee3\u7406\u5de5\u5177\u672c\u8eab\u5df2\u5f88\u64c5\u9577\u5354\u8abf\u6d41\u7a0b\uff0c\u4f8b\u5982\u7ba1\u7406\u5c0d\u8a71\u3001\u4e32\u63a5\u5de5\u5177\u548c\u57f7\u884c\u7a0b\u5f0f\u78bc\uff1b\u4f46\u4e00\u5230\u9700\u8981\u6574\u5408\u591a\u4efd\u6587\u4ef6\u3001\u5f15\u7528\u4f86\u6e90\u3001\u6216\u6839\u64da\u4f01\u696d\u8cc7\u6599\u6574\u7406\u9577\u7bc7\u5206\u6790\uff0c\u958b\u767c\u5718\u968a\u5f80\u5f80\u8981\u81ea\u884c\u8655\u7406\u5927\u91cf\u7e41\u7463\u5de5\u4f5c\u3002\u9019\u7bc7\u5167\u5bb9\u4ecb\u7d39\u7684 NVIDIA AI-Q\uff0c\u91cd\u9ede\u6b63\u662f\u628a\u9019\u985e\u6df1\u5ea6\u7814\u7a76\u80fd\u529b\u5c01\u88dd\u6210\u53ef\u651c\u7684\u9805\u76ee\u6280\u80fd\uff0c\u8b93\u4ee3\u7406\u5de5\u5177\u76f4\u63a5\u8abf\u7528\u3002<\/p>\n\n\n\n<p>\u5b83\u89e3\u6c7a\u7684\u554f\u984c\u5f88\u660e\u78ba\uff1a\u5718\u968a\u4e0d\u7528\u518d\u70ba\u6bcf\u500b\u9805\u76ee\u91cd\u5efa\u6aa2\u7d22\u3001\u898f\u5283\u3001\u6574\u7406\u3001\u5f15\u7528\u8207\u8a55\u4f30\u6d41\u7a0b\u3002\u4ee3\u7406\u5de5\u5177\u53ef\u628a\u7814\u7a76\u4efb\u52d9\u4ea4\u7d66\u672c\u6a5f\u6216\u8a17\u7ba1\u7684 AI-Q \u4f3a\u670d\u5668\uff0c\u4e4b\u5f8c\u6536\u56de\u4e00\u4efd\u6709\u7d50\u69cb\u3001\u9644\u5f15\u7528\u7684\u5831\u544a\uff1b\u800c\u654f\u611f\u8cc7\u6599\u4ea6\u53ef\u7559\u5728\u4f01\u696d\u5167\u90e8\u74b0\u5883\uff0c\u5c0d\u91ab\u7642\u3001\u91d1\u878d\u3001\u653f\u5e9c\u53ca\u570b\u9632\u7b49\u91cd\u8996\u8cc7\u6599\u7ba1\u63a7\u7684\u5834\u666f\u5c24\u5176\u91cd\u8981\u3002<\/p>\n\n\n\n<p>\u4e0a\u624b\u65b9\u5f0f\u4ea6\u76f8\u5c0d\u76f4\u63a5\u3002\u9019\u500b\u9805\u76ee\u63d0\u4f9b SKILL.md \u8aaa\u660e\u6a94\uff0c\u4ee5\u53ca\u4e00\u500b\u8f14\u52a9\u8173\u672c\u53bb\u8655\u7406\u8acb\u6c42\u8def\u7531\u3001\u63d0\u4ea4\u5de5\u4f5c\u3001\u8f2a\u8a62\u9032\u5ea6\u548c\u53d6\u56de\u7d50\u679c\u3002\u4f7f\u7528\u524d\u9700\u8981 Python 3.10 \u6216\u4ee5\u4e0a\uff0c\u4e26\u6e96\u5099\u4e00\u500b\u53ef\u9023\u7dda\u7684 AI-Q Blueprint \u4f3a\u670d\u5668\uff1b\u9810\u8a2d\u6703\u9023\u5230\u672c\u6a5f\u4f4d\u5740\uff0c\u4e5f\u53ef\u900f\u904e\u74b0\u5883\u8b8a\u6578\u6539\u5beb\u3002<\/p>\n\n\n<figure class=\"wp-block-embed-youtube wp-block-embed is-type-video is-provider-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"lyte-wrapper\" title=\"Give Codex a Deep Research Skill With NVIDIA AI-Q\" style=\"width:853px;max-width:100%;margin:5px auto;\"><div class=\"lyMe\" id=\"WYL_KNd7T3MgqZc\" itemprop=\"video\" itemscope itemtype=\"https:\/\/schema.org\/VideoObject\"><div><meta itemprop=\"thumbnailUrl\" content=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FKNd7T3MgqZc%2Fhqdefault.jpg\" \/><meta itemprop=\"embedURL\" content=\"https:\/\/www.youtube.com\/embed\/KNd7T3MgqZc\" \/><meta itemprop=\"duration\" content=\"PT4M11S\" \/><meta itemprop=\"uploadDate\" content=\"2026-05-20T16:09:21Z\" \/><\/div><div id=\"lyte_KNd7T3MgqZc\" data-src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FKNd7T3MgqZc%2Fhqdefault.jpg\" class=\"pL\"><div class=\"tC\"><div class=\"tT\" itemprop=\"name\">Give Codex a Deep Research Skill With NVIDIA AI-Q<\/div><\/div><div class=\"play\"><\/div><div class=\"ctrl\"><div class=\"Lctrl\"><\/div><div class=\"Rctrl\"><\/div><\/div><\/div><noscript><a href=\"https:\/\/youtu.be\/KNd7T3MgqZc\" rel=\"nofollow\"><img loading=\"lazy\" decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/plugins\/wp-youtube-lyte\/lyteCache.php?origThumbUrl=https%3A%2F%2Fi.ytimg.com%2Fvi%2FKNd7T3MgqZc%2F0.jpg\" alt=\"Give Codex a Deep Research Skill With NVIDIA AI-Q\" width=\"853\" height=\"460\" \/><br \/>Watch this video on YouTube<\/a><\/noscript><meta itemprop=\"description\" content=\"Developers building agent workflows often need their agents to research across multiple sources, synthesize findings, and return answers with citations. But, rebuilding that deep research pipeline inside every harness adds extra work. NVIDIA AI-Q gives agent harnesses like Codex a reusable way to delegate deep research instead of recreating retrieval, reasoning, synthesis, and citation logic in each agent. In this demo, Codex calls the AI-Q research skill, sends a multi-source research request to a running AI-Q server, follows the async checkpoint flow, and receives a structured report with citations. The walkthrough shows the basic architecture: Codex runs on the developer workstation, AI-Q handles the research workflow on a local or hosted server, and hosted NVIDIA inference models support planning, research, and embeddings. Use this pattern with Codex, Claude Code, OpenCode, or other agent harnesses to add grounded, source-attributed research to existing workflows. \u27a1\ufe0f Learn more: https:\/\/github.com\/NVIDIA-AI-Blueprints\/aiq#getting-started \ud83d\udcdd Tech blog: https:\/\/developer.nvidia.com\/blog\/add-a-specialized-deep-research-skill-to-agent-harnesses\/ \ud83d\udce5 Download: https:\/\/github.com\/NVIDIA-AI-Blueprints\/aiq\/tree\/develop\/.agents\/skills\/aiq-research 00:00 - Architecture Overview 00:30 - Install the AI-Q Research Skill 01:15 - Submit a Deep Research Query from Codex 02:05 - AI-Q Runs the Async Research Workflow 03:15 - Review the Structured Report and Citations 04:05 - Execution Trace and Skill Latency\"><\/div><\/div><div class=\"lL\" style=\"max-width:100%;width:853px;margin:5px auto;\"><\/div><figcaption><\/figcaption><\/figure>\n\n\n<p>\u91cd\u9ede\u53ef\u6982\u62ec\u5982\u4e0b\uff1a<br>&#8211; \u628a\u6df1\u5ea6\u7814\u7a76\u6d41\u7a0b\u5305\u88dd\u6210\u4ee3\u7406\u5de5\u5177\u53ef\u8abf\u7528\u7684\u6280\u80fd<br>&#8211; \u652f\u63f4\u591a\u4f86\u6e90\u8cc7\u6599\u6574\u5408\uff0c\u4e26\u8f38\u51fa\u9644\u5f15\u7528\u7684\u5831\u544a<br>&#8211; \u6e1b\u5c11\u6bcf\u500b\u9805\u76ee\u91cd\u8907\u642d\u5efa\u7814\u7a76\u6d41\u7a0b\u7684\u5de5\u4f5c<br>&#8211; \u8b93\u654f\u611f\u8cc7\u6599\u53ef\u4fdd\u7559\u5728\u4f01\u696d\u5167\u90e8\u74b0\u5883<br>&#8211; \u53ef\u914d\u5408 Claude Code\u3001Codex \u7b49\u901a\u7528\u4ee3\u7406\u5de5\u5177\u4f7f\u7528<\/p>\n\n\n\n<p>\u9019\u500b\u9805\u76ee\u7684\u5275\u65b0\u4e4b\u8655\uff0c\u5728\u65bc\u5b83\u4e0d\u662f\u53ea\u63d0\u4f9b\u55ae\u4e00\u5de5\u5177\u51fd\u5f0f\uff0c\u800c\u662f\u628a\u610f\u5716\u5206\u985e\u3001\u6f84\u6e05\u554f\u984c\u3001\u6dfa\u5c64\u7814\u7a76\u3001\u6df1\u5ea6\u7814\u7a76\u8207\u8a55\u4f30\u6574\u5408\u6210\u8f03\u9ad8\u5c64\u7684\u80fd\u529b\u3002\u63db\u53e5\u8a71\u8aaa\uff0c\u4ee3\u7406\u5de5\u5177\u53ea\u9700\u5b78\u6703\u5982\u4f55\u59d4\u6d3e\uff0c\u4fbf\u53ef\u5229\u7528\u5b8c\u6574\u7814\u7a76\u7ba1\u7dda\uff0c\u8f03\u9069\u5408\u60f3\u5feb\u901f\u70ba\u73fe\u6709\u4ee3\u7406\u7cfb\u7d71\u52a0\u5165\u7814\u7a76\u80fd\u529b\u7684\u5718\u968a\u3002\u6587\u4e2d\u672a\u898b\u5177\u9ad4\u6548\u80fd\u6578\u5b57\u6216\u57fa\u6e96\u6e2c\u8a66\uff0c\u56e0\u6b64\u73fe\u968e\u6bb5\u8f03\u9069\u5408\u628a\u5b83\u7406\u89e3\u70ba\u4e00\u500b\u5de5\u7a0b\u6574\u5408\u8207\u5de5\u4f5c\u6d41\u7a0b\u7c21\u5316\u65b9\u6848\u3002<\/p>\n\n\n\n<p>\u6a21\u578b\u5217\u8868\uff1aClaude Code\u3001Codex<\/p>\n\n\n\n<p><strong>GitHub\uff1a<\/strong> <a href=\"https:\/\/github.com\/NVIDIA-AI-Blueprints\/aiq\/tree\/v2.1.0\" rel=\"noopener noreferrer\">https:\/\/github.com\/NVIDIA-AI-Blueprints\/aiq\/tree\/v2.1.0<\/a><\/p>\n\n\n\n<p><strong>\u9805\u76ee\uff1a<\/strong> <a href=\"https:\/\/developer.nvidia.com\/blog\/add-a-specialized-deep-research-skill-to-agent-harnesses\/\" rel=\"noopener noreferrer\">https:\/\/developer.nvidia.com\/blog\/add-a-specialized-deep-research-skill-to-agent-harnesses\/<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>NVIDIA AI-Q\u628a\u6df1\u5ea6\u7814\u7a76\u6d41\u7a0b\u5c01\u88dd\u6210\u53ef\u91cd\u7528\u6280\u80fd\uff0c\u8b93\u4ee3\u7406\u5de5\u5177\u66f4\u6613\u8655\u7406\u591a\u4f86\u6e90\u5206\u6790\u3002<\/p>\n","protected":false},"author":8,"featured_media":8497,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[133,179,116,197],"tags":[],"class_list":["post-8498","post","type-post","status-publish","format-standard","hentry","category-133","category-nvidia","category-agentic","category-framework"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/8498","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/comments?post=8498"}],"version-history":[{"count":1,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/8498\/revisions"}],"predecessor-version":[{"id":8499,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/8498\/revisions\/8499"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media\/8497"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=8498"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/categories?post=8498"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/tags?post=8498"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}