{"id":115171,"date":"2026-10-11T12:08:55","date_gmt":"2026-10-11T04:08:55","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/115171.html"},"modified":"2026-10-11T12:08:55","modified_gmt":"2026-10-11T04:08:55","slug":"%e7%ac%ac1%e7%ab%a0%ef%bc%9acrewai-%e6%9c%af%e8%af%ad%e5%85%a8%e6%99%af%e4%b8%8e%e5%a4%9a%e6%99%ba%e8%83%bd%e4%bd%93%e5%b7%a5%e4%bd%9c%e5%8e%9f%e7%90%86","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/115171.html","title":{"rendered":"\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406"},"content":{"rendered":"

1. \u9879\u76ee\u80cc\u666f<\/h3>\n

\u5982\u679c\u4f60\u662f\u4e00\u5bb6\u4e2d\u578b\u7535\u5546\u516c\u53f8\u7684\u6280\u672f\u8d1f\u8d23\u4eba，\u4f60\u4f1a\u53d1\u73b0\u56e2\u961f\u6b63\u5728\u7ecf\u5386\u4e00\u573a"AI \u7126\u8651"\u3002\u8fd0\u8425\u90e8\u95e8\u5e0c\u671b\u7528\u5927\u6a21\u578b\u81ea\u52a8\u751f\u6210\u5546\u54c1\u6587\u6848，\u5ba2\u670d\u56e2\u961f\u60f3\u8981\u4e00\u4e2a\u80fd\u56de\u7b54 80% \u5e38\u89c1\u95ee\u9898\u7684\u673a\u5668\u4eba，\u6570\u636e\u5206\u6790\u5e08\u5219\u60f3\u628a\u6bcf\u5468\u4e94\u7684\u62a5\u8868\u4ece"\u624b\u52a8\u62d6\u6570\u2014\u2014Excel \u753b\u56fe\u2014\u2014\u5199\u7ed3\u8bba"\u538b\u7f29\u5230\u4e00\u952e\u751f\u6210\u3002<\/p>\n

\u4e8e\u662f\u4f60\u8ba9\u56e2\u961f\u5404\u81ea\u5c1d\u8bd5：\u8fd0\u8425\u7528 GPT-4 \u5199 prompt \u751f\u6210\u6587\u6848，\u7ed3\u679c\u8d28\u91cf\u5ffd\u9ad8\u5ffd\u4f4e\u2014\u2014\u540c\u6837\u7684 prompt，\u6709\u65f6\u8f93\u51fa\u60ca\u8273，\u6709\u65f6\u5b8c\u5168\u8dd1\u504f；\u5ba2\u670d\u7528 LangChain \u642d\u4e86\u4e00\u4e2a\u95ee\u7b54\u94fe，\u53d1\u73b0\u4e00\u65e6\u7528\u6237\u7684\u95ee\u9898\u6d89\u53ca\u8ba2\u5355\u67e5\u8be2\u3001\u9000\u6362\u8d27\u72b6\u6001，\u6a21\u578b\u5c31\u5f00\u59cb"\u4e00\u672c\u6b63\u7ecf\u5730\u80e1\u8bf4\u516b\u9053"；\u6570\u636e\u5206\u6790\u5e08\u7528 Python \u811a\u672c\u8c03 API，\u6bcf\u6b21\u90fd\u8981\u624b\u52a8\u62fc\u63a5\u4e0a\u4e0b\u6587，\u50cf\u5728\u6307\u6325\u4e00\u4e2a"\u53ea\u4f1a\u6267\u884c\u3001\u4e0d\u4f1a\u601d\u8003"\u7684\u5de5\u5177\u4eba\u3002<\/p>\n

\u6838\u5fc3\u75db\u70b9：\u5355\u4e00\u5927\u6a21\u578b\u53ea\u80fd\u5728"\u7ed9\u5b9a\u4e0a\u4e0b\u6587 + \u7ed9\u5b9a\u95ee\u9898"\u7684\u8303\u56f4\u5185\u56de\u7b54，\u4f46\u771f\u5b9e\u4e1a\u52a1\u5f80\u5f80\u662f"\u591a\u6b65\u9aa4\u63a8\u7406 + \u591a\u89d2\u8272\u534f\u4f5c + \u591a\u5de5\u5177\u8c03\u7528"\u7684\u590d\u5408\u573a\u666f\u3002\u6bd4\u5982"\u5e2e\u6211\u5206\u6790\u7ade\u54c1\u5b9a\u4ef7\u7b56\u7565"\u8fd9\u4e2a\u9700\u6c42，\u80cc\u540e\u9700\u8981：\u4fe1\u606f\u68c0\u7d22\u5458\u53bb\u641c\u7d22\u7ade\u54c1\u5b98\u7f51\u548c\u65b0\u95fb \u2192 \u6570\u636e\u5206\u6790\u5e08\u6574\u7406\u4ef7\u683c\u533a\u95f4 \u2192 \u7b56\u7565\u5206\u6790\u5e08\u8f93\u51fa\u7ed3\u8bba\u548c\u98ce\u9669\u5efa\u8bae\u3002\u8fd9\u5df2\u7ecf\u4e0d\u662f\u5355\u4e2a Agent \u80fd\u505a\u7684\u4e8b\u3002<\/p>\n

\u4f60\u7f3a\u7684\u4e0d\u662f\u5927\u6a21\u578b\u7684\u80fd\u529b，\u800c\u662f\u4e00\u4e2a\u80fd\u628a\u5927\u6a21\u578b\u80fd\u529b\u7f16\u6392\u6210\u56e2\u961f\u534f\u4f5c\u7684\u6846\u67b6\u2014\u2014\u8fd9\u5c31\u662f CrewAI \u8981\u89e3\u51b3\u7684\u95ee\u9898\u3002<\/p>\n

2. \u9879\u76ee\u8bbe\u8ba1<\/h3>\n

\u5c0f\u80d6（\u5634\u91cc\u585e\u7740\u85af\u7247）：\u201c\u5927\u5e08，\u8fd9\u4e2a CrewAI \u5230\u5e95\u662f\u4e2a\u5565？\u6211\u542c\u540d\u5b57\u8fd8\u4ee5\u4e3a\u8ddf\u5e06\u8239\u8fd0\u52a8\u6709\u5173\u3002\u201d<\/p>\n

\u5927\u5e08（\u63a8\u4e86\u63a8\u773c\u955c）：\u201c\u4f60\u53ef\u4ee5\u628a\u5b83\u60f3\u8c61\u6210\u4e00\u4e2a\u516c\u53f8\u3002Crew \u662f\u516c\u53f8，Agent \u662f\u4e0d\u540c\u90e8\u95e8\u7684\u5458\u5de5，Task \u662f\u6bcf\u4e2a\u4eba\u624b\u4e2d\u7684\u5de5\u4f5c\u6e05\u5355，Tool \u662f\u5458\u5de5\u684c\u4e0a\u7684\u7535\u8bdd\u548c\u7535\u8111\u3002\u800c Process，\u5c31\u662f\u516c\u53f8\u5185\u90e8\u7684\u5de5\u4f5c\u6d41\u7a0b\u2014\u2014\u662f\u6309\u5de5\u4f4d\u987a\u5e8f\u4f20\u9012，\u8fd8\u662f\u8bbe\u4e2a\u7ecf\u7406\u6765\u5206\u914d\u4efb\u52a1\u3002\u201d<\/p>\n

\u5c0f\u80d6：\u201c\u54e6\u2026\u2026\u90a3\u4e0d\u5c31\u662f\u5de5\u4f5c\u6d41\u5f15\u64ce\u5417？\u8ddf Jenkins pipeline \u5dee\u4e0d\u591a？\u201d<\/p>\n

\u5c0f\u767d（\u653e\u4e0b\u4e66）：\u201c\u4e0d\u5bf9\u3002Jenkins \u662f\u786e\u5b9a\u6027\u6d41\u7a0b\u2014\u2014\u6bcf\u4e2a\u6b65\u9aa4\u7684\u8f93\u5165\u8f93\u51fa\u90fd\u662f\u56fa\u5b9a\u7684\u3002\u4f46 CrewAI \u91cc\u7684 Agent \u662f\u8ba9\u5927\u6a21\u578b\u6765\u2019\u601d\u8003\u2019\u4e0b\u4e00\u6b65\u8be5\u505a\u4ec0\u4e48，\u8f93\u51fa\u662f\u4e0d\u786e\u5b9a\u7684\u3002\u8fd9\u4e2d\u95f4\u600e\u4e48\u4fdd\u8bc1\u8d28\u91cf？\u201d<\/p>\n

\u5927\u5e08：\u201c\u95ee\u5f97\u597d\u3002\u8fd9\u5c31\u662f Guardrail \u8981\u89e3\u51b3\u7684\u95ee\u9898\u3002\u5de5\u4f5c\u6d41\u5f15\u64ce\u7684\u6bcf\u4e00\u6b65\u7ed3\u679c\u662f\u786e\u5b9a\u7684\u4e8c\u8fdb\u5236\u2014\u2014\u6210\u529f\u6216\u5931\u8d25\u3002\u4f46 Agent \u7684\u8f93\u51fa\u53ef\u80fd\u662f\u2019\u770b\u8d77\u6765\u5bf9\u4f46\u5b9e\u9645\u6709\u8bef\u2019\u7684，\u6240\u4ee5\u9700\u8981 Guardrail \u6765\u505a\u683c\u5f0f\u6821\u9a8c\u3001\u4e1a\u52a1\u89c4\u5219\u6821\u9a8c\u3001\u5b89\u5168\u6821\u9a8c\u3002\u5982\u679c\u6821\u9a8c\u4e0d\u901a\u8fc7，\u89e6\u53d1\u91cd\u8bd5\u6216\u4eba\u5de5\u63a5\u7ba1\u3002\u201d<\/p>\n

\u5c0f\u80d6：\u201c\u90a3 Memory \u53c8\u662f\u4ec0\u4e48？\u76f8\u5f53\u4e8e\u5458\u5de5\u7684\u5de5\u4f5c\u7b14\u8bb0\u672c\u5417？\u201d<\/p>\n

\u5927\u5e08：\u201c\u5dee\u4e0d\u591a\u3002Memory \u5206\u6210\u4e09\u5c42\u2014\u2014\u77ed\u671f\u8bb0\u5fc6\u8bb0\u5f55\u5f53\u524d\u5bf9\u8bdd\u7684\u4e0a\u4e0b\u6587，\u957f\u671f\u8bb0\u5fc6\u4fdd\u5b58\u7528\u6237\u7684\u504f\u597d\u548c\u5386\u53f2\u6a21\u5f0f，\u5b9e\u4f53\u8bb0\u5fc6\u8bb0\u5f55\u50cf\u2019\u5f20\u4e09\u662f VIP \u5ba2\u6237\u2019\u8fd9\u6837\u7684\u5173\u7cfb\u3002\u4f46\u8981\u6ce8\u610f，\u8bb0\u5fc6\u591a\u4e86\u4f1a\u6c61\u67d3\u2014\u2014\u5c31\u50cf\u7b14\u8bb0\u672c\u4e0a\u8bb0\u7684\u4e1c\u897f\u592a\u591a，\u7ffb\u56de\u53bb\u627e\u8981\u82b1\u65f6\u95f4，\u8fd8\u53ef\u80fd\u8bb0\u9519\u3002\u201d<\/p>\n

\u5c0f\u767d：\u201c\u6211\u8fd8\u6709\u4e00\u4e2a\u7591\u95ee\u3002Knowledge \u548c Memory \u7684\u533a\u522b\u5728\u54ea？\u542c\u8d77\u6765\u90fd\u662f\u2019\u5b58\u4fe1\u606f\u2019\u3002\u201d<\/p>\n

\u5927\u5e08：\u201c\u5173\u952e\u533a\u522b\u5728\u4e8e\u4fe1\u606f\u6765\u6e90\u3002Memory \u662f\u6211\u6267\u884c\u4efb\u52a1\u8fc7\u7a0b\u4e2d\u79ef\u7d2f\u7684\u7ecf\u9a8c；Knowledge \u662f\u5916\u90e8\u77e5\u8bc6\u2014\u2014\u516c\u53f8\u7684\u5236\u5ea6\u6587\u6863\u3001\u4ea7\u54c1\u624b\u518c\u3001\u6280\u672f\u89c4\u8303\u3002Memory \u5e2e\u4f60\u8bb0\u4f4f\u2019\u4e0a\u6b21\u8fd9\u4e48\u5904\u7406\u78b0\u4e86\u58c1\u2019，Knowledge \u5e2e\u4f60\u67e5\u5230\u2019\u516c\u53f8\u89c4\u5b9a\u9000\u6b3e\u65f6\u6548\u662f 7 \u5929\u2019\u3002\u4e24\u8005\u7ed3\u5408，Agent \u624d\u80fd\u65e2\u806a\u660e\u53c8\u51c6\u786e\u3002\u201d<\/p>\n

\u5c0f\u80d6：\u201c\u90a3 Flow \u53c8\u662f\u4ec0\u4e48？\u542c\u8d77\u6765\u8ddf Crew \u5dee\u4e0d\u591a\u554a\u3002\u201d<\/p>\n

\u5927\u5e08：\u201cCrew \u662f\u8ba9 Agent \u4eec\u534f\u4f5c\u5b8c\u6210\u4e00\u4e2a\u76ee\u6807，\u63a7\u5236\u6743\u5728 Agent \u624b\u91cc\u2014\u2014\u8c01\u5148\u505a\u3001\u8c01\u540e\u505a，Process \u5b9a\u4e49\u4e86\u57fa\u8c03\u3002Flow \u4e0d\u4e00\u6837\u2014\u2014Flow \u662f\u4f60\u7528\u786e\u5b9a\u6027\u7684 Python \u4ee3\u7801\u6765\u63a7\u5236\u6d41\u7a0b\u8d70\u5411\u3002\u6bd4\u5982\u2019\u8bc6\u522b\u5de5\u5355\u7c7b\u578b \u2192 \u5982\u679c\u662f\u9000\u6b3e，\u8d70\u9000\u6b3e\u5206\u652f；\u5982\u679c\u662f\u6295\u8bc9，\u8d70\u5347\u7ea7\u5904\u7406\u5206\u652f\u2019\u2014\u2014\u8fd9\u79cd\u5206\u652f\u8def\u7531\u903b\u8f91\u7528 Flow \u6765\u63a7\u5236，\u6bd4\u8ba9 Agent \u81ea\u5df1\u5224\u65ad\u66f4\u53ef\u9760\u3002\u201d<\/p>\n

\u5c0f\u767d：\u201c\u6240\u4ee5 Flow \u662f\u2019\u786e\u5b9a\u6027\u6d41\u7a0b\u2019，Crew \u662f\u2019\u667a\u80fd\u534f\u4f5c\u2019，\u4e24\u8005\u53ef\u4ee5\u5d4c\u5957\u4f7f\u7528？\u201d<\/p>\n

\u5927\u5e08：\u201cExactly\u3002\u4f60\u53ef\u4ee5\u628a Crew \u5d4c\u5165\u5230 Flow \u7684\u4e00\u4e2a\u8282\u70b9\u91cc，\u4e5f\u53ef\u4ee5\u8ba9 Flow \u4e2d\u7684\u67d0\u4e2a\u6b65\u9aa4\u89e6\u53d1\u4e00\u4e2a Crew \u6267\u884c\u3002\u8fd9\u662f CrewAI \u6700\u7075\u6d3b\u7684\u5730\u65b9\u3002\u201d<\/p>\n

\u6280\u672f\u6620\u5c04\u603b\u7ed3：Agent \u662f\u6267\u884c\u5355\u5143，Task \u662f\u76ee\u6807\u5355\u5143，Tool \u662f\u80fd\u529b\u5355\u5143，Crew \u662f\u7ec4\u7ec7\u5355\u5143，Process \u662f\u534f\u4f5c\u6a21\u5f0f，Flow \u662f\u63a7\u5236\u6d41\u7a0b，Memory \u662f\u7ecf\u9a8c\u5b58\u50a8，Knowledge \u662f\u5916\u90e8\u77e5\u8bc6，Guardrail \u662f\u8d28\u91cf\u5173\u53e3，LLM \u662f\u63a8\u7406\u5f15\u64ce\u3002\u7406\u89e3\u8fd9\u5341\u4e2a\u672f\u8bed\u7684\u804c\u8d23\u5206\u5de5，\u5c31\u7406\u89e3\u4e86 CrewAI \u7684\u6838\u5fc3\u67b6\u6784\u3002<\/p>\n

\u67b6\u6784\u56fe<\/h4>\n

\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510
\n\u2502 \u7528\u6237\u76ee\u6807 \/ \u4e1a\u52a1\u9700\u6c42 \u2502
\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518
\n \u2502
\n \u25bc
\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510
\n\u2502 Crew \/ Flow（\u7f16\u6392\u5c42） \u2502
\n\u2502 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u2502
\n\u2502 \u2502 Sequential\u2502 \u2502Hierarchical\u2502 \u2502 Flow \u2502 \u2502 Flow + Crew \u2502 \u2502
\n\u2502 \u2502 Process \u2502 \u2502 Process \u2502 \u2502 (Router) \u2502 \u2502 (\u5d4c\u5957\u7ec4\u5408) \u2502 \u2502
\n\u2502 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2502
\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518
\n \u2502
\n \u25bc
\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510
\n\u2502 Agent \u667a\u80fd\u4f53\u5c42 \u2502
\n\u2502 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u2502
\n\u2502 \u2502 \u7814\u7a76\u5458 Agent \u2502 \u2502 \u5206\u6790\u5e08 Agent \u2502 \u2502 \u5ba1\u6838\u5458 Agent \u2502 … \u2502
\n\u2502 \u2502 role\/goal\/ \u2502 \u2502 role\/goal\/ \u2502 \u2502 role\/goal\/ \u2502 \u2502
\n\u2502 \u2502 backstory \u2502 \u2502 backstory \u2502 \u2502 backstory \u2502 \u2502
\n\u2502 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u252c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u252c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u252c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2502
\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u253c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u253c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u253c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518
\n \u2502 \u2502 \u2502
\n \u25bc \u25bc \u25bc
\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510
\n\u2502 Task \u4efb\u52a1 + Context \u4e0a\u4e0b\u6587 \u2502
\n\u2502 Task 1 \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500>
\n\u2502 \u2192 Task 2 \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500>
\n\u2502 \u2192 Task 3 \u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500>
\n\u2502 { context: \u524d\u5e8f\u4efb\u52a1\u7684\u8f93\u51fa } { expected_output: \u7ed3\u6784\u5316Schema }
\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518
\n \u2502
\n \u25bc
\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510
\n\u2502 Tool \u5de5\u5177\u8c03\u7528\u5c42 \u2502
\n\u2502 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u2502
\n\u2502 \u2502 WebSearch \u2502 \u2502 FileRead \u2502 \u2502 APICall \u2502 \u2502 DBQuery \u2502 \u2502 Custom \u2502 \u2502
\n\u2502 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2502
\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518
\n \u2502
\n \u25bc
\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510
\n\u2502 Memory + Knowledge \u5b58\u50a8\u5c42 \u2502
\n\u2502 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u2502
\n\u2502 \u2502 Short-term Mem \u2502 \u2502 Long-term Mem \u2502 \u2502 Entity Memory \u2502 \u2502
\n\u2502 \u2502 (\u5f53\u524d\u5bf9\u8bdd\u4e0a\u4e0b\u6587) \u2502 \u2502 (\u7528\u6237\u504f\u597d\/\u5386\u53f2) \u2502 \u2502 (\u5b9e\u4f53\u5173\u7cfb) \u2502 \u2502
\n\u2502 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2502
\n\u2502 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510\u2502
\n\u2502 \u2502 Knowledge (RAG): \u6587\u6863 \u2192 \u5207\u5206 \u2192 \u5411\u91cf\u5316 \u2192 \u68c0\u7d22 \u2192 \u589e\u5f3a\u751f\u6210 \u2502\u2502
\n\u2502 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518\u2502
\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518
\n \u2502
\n \u25bc
\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510
\n\u2502 LLM Provider \u6a21\u578b\u5c42 \u2502
\n\u2502 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u2502
\n\u2502 \u2502 OpenAI \u2502 \u2502 Azure OpenAI \u2502 \u2502 Ollama \u2502 … \u2502
\n\u2502 \u2502 (GPT-4o) \u2502 \u2502 (GPT-4\/3.5) \u2502 \u2502 (\u672c\u5730\u6a21\u578b) \u2502 \u2502
\n\u2502 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2502
\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518
\n \u2502
\n \u25bc
\n\u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510
\n\u2502 Guardrail \u8d28\u91cf\u6821\u9a8c\u5c42 \u2502
\n\u2502 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u250c\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2510 \u2502
\n\u2502 \u2502 \u683c\u5f0f\u6821\u9a8c \u2502 \u2502 \u5b57\u6bb5\u5b8c\u6574\u6027\u2502 \u2502 \u4e1a\u52a1\u89c4\u5219 \u2502 \u2502 \u4eba\u5de5\u786e\u8ba4(Human-in- \u2502 \u2502
\n\u2502 \u2502 \u2502 \u2502 \u2502 \u2502 \u2502 \u2502 the-Loop) \u2502 \u2502
\n\u2502 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518 \u2502
\n\u2514\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2500\u2518<\/p>\n

CrewAI \u6838\u5fc3\u5de5\u4f5c\u94fe\u8def：<\/p>\n

\u7528\u6237\u76ee\u6807
\n \u2192 (Flow\/Crew) \u4efb\u52a1\u62c6\u89e3
\n \u2192 (Agent) \u89d2\u8272\u5339\u914d\u4e0e\u63a8\u7406
\n \u2192 (Tool) \u5916\u90e8\u80fd\u529b\u8c03\u7528
\n \u2192 (LLM) \u6a21\u578b\u63a8\u7406
\n \u2192 (Memory\/Knowledge) \u4e0a\u4e0b\u6587\u589e\u5f3a
\n \u2192 (Guardrail) \u8d28\u91cf\u6821\u9a8c
\n \u2192 \u7ed3\u679c\u6c47\u603b\u4e0e\u8f93\u51fa<\/p>\n

\u4e09\u79cd\u7f16\u6392\u601d\u8def\u7684\u6838\u5fc3\u533a\u522b：<\/p>\n\n\u7ef4\u5ea6Sequential ProcessHierarchical ProcessFlow<\/tr>\n\n\n\n\n\n
\u63a7\u5236\u65b9\u5f0f<\/td>\n\u6309 Task \u5217\u8868\u987a\u5e8f\u6267\u884c<\/td>\nManager Agent \u52a8\u6001\u5206\u914d<\/td>\nPython \u4ee3\u7801\u786e\u5b9a\u6027\u8def\u7531<\/td>\n<\/tr>\n
\u9002\u7528\u573a\u666f<\/td>\n\u7ebf\u6027\u6d41\u6c34\u7ebf（\u5206\u6790\u2192\u5199\u7a3f\u2192\u5ba1\u6838）<\/td>\n\u590d\u6742\u534f\u4f5c，\u4efb\u52a1\u4e0d\u786e\u5b9a（\u591a\u8f6e\u8ba8\u8bba\u8bc4\u5ba1）<\/td>\n\u6761\u4ef6\u5206\u652f\u591a\u3001\u9700\u4e0e\u5916\u90e8\u7cfb\u7edf\u4ea4\u4e92<\/td>\n<\/tr>\n
\u7075\u6d3b\u6027<\/td>\n\u4f4e，\u4f46\u53ef\u9884\u6d4b<\/td>\n\u9ad8，\u4f46\u9700\u6ce8\u610f Manager \u5931\u63a7<\/td>\n\u6700\u9ad8，\u53ef\u5d4c\u5165 Crew<\/td>\n<\/tr>\n
\u590d\u6742\u5ea6<\/td>\n\u7b80\u5355<\/td>\n\u4e2d\u7b49<\/td>\n\u8f83\u9ad8<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

3. \u9879\u76ee\u5b9e\u6218<\/h3>\n

3.1 \u5b9e\u6218\u76ee\u6807<\/h4>\n

\u7528 Mermaid \u548c Python \u4ee3\u7801，\u753b\u51fa"AI \u5e02\u573a\u8c03\u7814\u5c0f\u961f"\u7684 CrewAI \u67b6\u6784\u56fe，\u5e76\u8fd0\u884c\u4e00\u4e2a\u6700\u5c0f\u5316\u9a8c\u8bc1\u811a\u672c\u6765\u7406\u89e3\u5404\u7ec4\u4ef6\u7684\u5173\u7cfb\u3002<\/p>\n

3.2 \u73af\u5883\u51c6\u5907<\/h4>\n

# Python \u7248\u672c\u8981\u6c42：3.10 ~ 3.13<\/span>
\npython –version<\/span><\/p>\n

# \u521b\u5efa\u865a\u62df\u73af\u5883<\/span>
\npython -m<\/span> venv crewai-env<\/p>\n

# \u6fc0\u6d3b\u865a\u62df\u73af\u5883（Windows）<\/span>
\ncrewai-env\\\\<\/span>Scripts\\\\<\/span>activate<\/p>\n

# \u6fc0\u6d3b\u865a\u62df\u73af\u5883（macOS\/Linux）<\/span>
\nsource<\/span> crewai-env\/bin\/activate<\/p>\n

# \u5b89\u88c5\u4f9d\u8d56<\/span>
\npip install<\/span> crewai crewai-tools python-dotenv<\/p>\n

3.3 \u5206\u6b65\u5b9e\u73b0<\/h4>\n

\u6b65\u9aa41：\u521b\u5efa\u9879\u76ee\u7ed3\u6784<\/h5>\n

\u76ee\u6807：\u5efa\u7acb\u6807\u51c6\u5316\u7684 CrewAI \u9879\u76ee\u76ee\u5f55\u3002<\/p>\n

market-research-crew\/
\n\u251c\u2500\u2500 .env # API Key \u914d\u7f6e
\n\u251c\u2500\u2500 main.py # \u4e3b\u5165\u53e3
\n\u251c\u2500\u2500 agents.py # Agent \u5b9a\u4e49
\n\u251c\u2500\u2500 tasks.py # Task \u5b9a\u4e49
\n\u251c\u2500\u2500 tools.py # \u81ea\u5b9a\u4e49 Tool
\n\u251c\u2500\u2500 crew.py # Crew \u7f16\u6392
\n\u2514\u2500\u2500 outputs\/ # \u8f93\u51fa\u76ee\u5f55<\/p>\n

\u6b65\u9aa42：\u914d\u7f6e LLM Provider<\/h5>\n

\u76ee\u6807：\u914d\u7f6e OpenAI \u517c\u5bb9\u7684 API \u63a5\u53e3\u3002<\/p>\n

# .env<\/span>
\nOPENAI_API_KEY=<\/span>sk–<\/span>your–<\/span>api–<\/span>key–<\/span>here
\nOPENAI_API_BASE=<\/span>https:<\/span>\/\/<\/span>api.<\/span>openai.<\/span>com\/<\/span>v1 # \u6216\u5176\u4ed6\u517c\u5bb9\u7aef\u70b9<\/span>
\nMODEL_NAME=<\/span>gpt–<\/span>4o–<\/span>mini # \u8c03\u8bd5\u9636\u6bb5\u63a8\u8350\u4f4e\u6210\u672c\u6a21\u578b<\/span><\/p>\n

# main.py (\u6700\u5c0f\u5316\u9a8c\u8bc1)<\/span>
\nimport<\/span> os
\nfrom<\/span> dotenv import<\/span> load_dotenv<\/p>\n

load_dotenv(<\/span>)<\/span><\/p>\n

print<\/span>(<\/span>f"API Key loaded: <\/span>{<\/span>'Yes'<\/span> if<\/span> os.<\/span>getenv(<\/span>'OPENAI_API_KEY'<\/span>)<\/span> else<\/span> 'No'<\/span>}<\/span><\/span>"<\/span><\/span>)<\/span>
\nprint<\/span>(<\/span>f"Model: <\/span>{<\/span>os.<\/span>getenv(<\/span>'MODEL_NAME'<\/span>,<\/span> 'gpt-4o-mini'<\/span>)<\/span>}<\/span><\/span>"<\/span><\/span>)<\/span><\/p>\n

\u8fd0\u884c\u7ed3\u679c：<\/p>\n

API Key loaded: Yes
\nModel: gpt-4o-mini<\/p>\n

\u5e38\u89c1\u5751：Windows \u7528\u6237\u53ef\u80fd\u4f1a\u9047\u5230\u7f16\u7801\u95ee\u9898\u3002\u5982\u679c\u51fa\u73b0 Unicode \u9519\u8bef，\u5728 .env \u6587\u4ef6\u4e2d\u786e\u4fdd\u4f7f\u7528 UTF-8 \u7f16\u7801\u4fdd\u5b58\u3002<\/p>\n

\u6b65\u9aa43：\u5b9a\u4e49\u4e09\u4e2a Agent<\/h5>\n

\u76ee\u6807：\u4e3a"AI \u5e02\u573a\u8c03\u7814\u5c0f\u961f"\u5b9a\u4e49\u4e09\u4e2a\u89d2\u8272\u3002<\/p>\n

# agents.py<\/span>
\nfrom<\/span> crewai import<\/span> Agent<\/p>\n

def<\/span> create_researcher<\/span>(<\/span>llm)<\/span>:<\/span>
\n """\u4fe1\u606f\u68c0\u7d22\u5458：\u8d1f\u8d23\u641c\u7d22\u548c\u6536\u96c6\u5e02\u573a\u4fe1\u606f"""<\/span>
\n return<\/span> Agent(<\/span>
\n role=<\/span>"\u5e02\u573a\u4fe1\u606f\u68c0\u7d22\u4e13\u5bb6"<\/span>,<\/span>
\n goal=<\/span>"\u641c\u7d22\u76ee\u6807\u5e02\u573a\u7684\u7ade\u54c1\u4fe1\u606f\u3001\u884c\u4e1a\u62a5\u544a\u548c\u6700\u65b0\u52a8\u6001，\u4e0d\u9057\u6f0f\u4efb\u4f55\u5173\u952e\u6570\u636e\u6e90"<\/span>,<\/span>
\n backstory=<\/span>(<\/span>
\n "\u4f60\u66fe\u5728\u9ea6\u80af\u9521\u62c5\u4efb\u884c\u4e1a\u7814\u7a76\u5458，\u64c5\u957f\u4ece\u6d77\u91cf\u4fe1\u606f\u4e2d\u5feb\u901f\u5b9a\u4f4d\u5173\u952e\u6570\u636e\u3002"<\/span>
\n "\u4f60\u4e60\u60ef\u5148\u5217\u68c0\u7d22\u7b56\u7565，\u518d\u9010\u4e00\u6536\u96c6，\u6700\u540e\u6309\u53ef\u9760\u6027\u6392\u5e8f\u3002"<\/span>
\n )<\/span>,<\/span>
\n llm=<\/span>llm,<\/span>
\n verbose=<\/span>True<\/span>,<\/span>
\n allow_delegation=<\/span>False<\/span> # \u7814\u7a76\u5458\u4e0d\u59d4\u6258\u4efb\u52a1，\u4e13\u6ce8\u68c0\u7d22<\/span>
\n )<\/span><\/p>\n

def<\/span> create_analyst<\/span>(<\/span>llm)<\/span>:<\/span>
\n """\u6570\u636e\u5206\u6790\u5e08：\u8d1f\u8d23\u5206\u6790\u6570\u636e\u5e76\u63d0\u70bc\u6d1e\u5bdf"""<\/span>
\n return<\/span> Agent(<\/span>
\n role=<\/span>"\u6570\u636e\u5206\u6790\u5e08"<\/span>,<\/span>
\n goal=<\/span>"\u5c06\u539f\u59cb\u5e02\u573a\u6570\u636e\u8f6c\u5316\u4e3a\u53ef\u843d\u5730\u6d1e\u5bdf，\u8f93\u51fa\u4ef7\u683c\u533a\u95f4\u3001\u5dee\u5f02\u5316\u5356\u70b9\u548c\u98ce\u9669\u9884\u5224"<\/span>,<\/span>
\n backstory=<\/span>(<\/span>
\n "\u4f60\u670910\u5e74\u4e92\u8054\u7f51\u884c\u4e1a\u6570\u636e\u5206\u6790\u7ecf\u9a8c，\u64c5\u957f\u4ece\u770b\u4f3c\u65e0\u5173\u7684\u6570\u636e\u4e2d\u53d1\u73b0\u5173\u8054\u6a21\u5f0f\u3002"<\/span>
\n "\u4f60\u5bf9\u6570\u5b57\u654f\u611f，\u4f46\u4ece\u4e0d\u6ee1\u8db3\u4e8e\u62a5\u8868\u2014\u2014\u4f60\u8ffd\u95ee\u80cc\u540e\u7684'\u4e3a\u4ec0\u4e48'\u3002"<\/span>
\n )<\/span>,<\/span>
\n llm=<\/span>llm,<\/span>
\n verbose=<\/span>True<\/span>,<\/span>
\n allow_delegation=<\/span>False<\/span>
\n )<\/span><\/p>\n

def<\/span> create_strategist<\/span>(<\/span>llm)<\/span>:<\/span>
\n """\u7b56\u7565\u5206\u6790\u5e08：\u8d1f\u8d23\u8f93\u51fa\u6700\u7ec8\u7ed3\u8bba\u548c\u884c\u52a8\u5efa\u8bae"""<\/span>
\n return<\/span> Agent(<\/span>
\n role=<\/span>"\u7b56\u7565\u5206\u6790\u5e08"<\/span>,<\/span>
\n goal=<\/span>"\u57fa\u4e8e\u6570\u636e\u5206\u6790\u7ed3\u679c，\u8f93\u51fa\u5b8c\u6574\u7684\u7ade\u4e89\u7b56\u7565\u5efa\u8bae\u548c\u6267\u884c\u8def\u7ebf\u56fe"<\/span>,<\/span>
\n backstory=<\/span>(<\/span>
\n "\u4f60\u662f\u4e00\u540d\u8d44\u6df1\u4ea7\u54c1\u6218\u7565\u987e\u95ee，\u4e3a\u8d85\u8fc750\u5bb6\u516c\u53f8\u5236\u5b9a\u8fc7\u5e02\u573a\u8fdb\u5165\u7b56\u7565\u3002"<\/span>
\n "\u4f60\u5173\u6ce8\u4e09\u4ef6\u4e8b：\u5bf9\u624b\u5728\u505a\u4ec0\u4e48\u3001\u6211\u4eec\u6709\u4ec0\u4e48\u4f18\u52bf\u3001\u63a5\u4e0b\u6765\u4e09\u4e2a\u6708\u8be5\u505a\u4ec0\u4e48\u3002"<\/span>
\n )<\/span>,<\/span>
\n llm=<\/span>llm,<\/span>
\n verbose=<\/span>True<\/span>,<\/span>
\n allow_delegation=<\/span>True<\/span> # \u7b56\u7565\u5e08\u53ef\u4ee5\u59d4\u6258\u4efb\u52a1\u7ed9\u5176\u4ed6 Agent<\/span>
\n )<\/span><\/p>\n

\u6b65\u9aa44：\u5b9a\u4e49\u4efb\u52a1\u94fe<\/h5>\n

\u76ee\u6807：\u628a"\u7ade\u54c1\u5206\u6790\u62a5\u544a"\u62c6\u6210\u53ef\u6267\u884c\u7684\u4e09\u6b65\u4efb\u52a1\u3002<\/p>\n

# tasks.py<\/span>
\nfrom<\/span> crewai import<\/span> Task<\/p>\n

def<\/span> create_research_task<\/span>(<\/span>agent,<\/span> product_name)<\/span>:<\/span>
\n """\u7b2c\u4e00\u6b65：\u6536\u96c6\u4fe1\u606f"""<\/span>
\n return<\/span> Task(<\/span>
\n description=<\/span>f"""
\n \u8bf7\u641c\u7d22"<\/span>{<\/span>product_name}<\/span><\/span>"\u76f8\u5173\u7684\u5e02\u573a\u4fe1\u606f，\u5305\u62ec：
\n 1. \u4e3b\u8981\u7ade\u54c1\u540d\u79f0\u3001\u5b9a\u4ef7\u3001\u6838\u5fc3\u529f\u80fd
\n 2. \u8fd1\u534a\u5e74\u7684\u884c\u4e1a\u62a5\u544a\u548c\u878d\u8d44\u52a8\u6001
\n 3. \u76ee\u6807\u7528\u6237\u7fa4\u4f53\u7684\u75db\u70b9\u548c\u672a\u6ee1\u8db3\u9700\u6c42<\/p>\n

\u6570\u636e\u6765\u6e90\u4f18\u5148\u7ea7：\u5b98\u65b9\u7f51\u7ad9 > \u884c\u4e1a\u62a5\u544a > \u79d1\u6280\u5a92\u4f53 > \u793e\u533a\u8ba8\u8bba
\n """<\/span><\/span>,<\/span>
\n expected_output=<\/span>(<\/span>
\n "\u4e00\u4efd\u7ed3\u6784\u5316\u7684\u5e02\u573a\u8c03\u7814\u62a5\u544a，\u5305\u542b：\\\\n"<\/span>
\n "- \u7ade\u54c1\u5217\u8868（\u540d\u79f0\u3001\u5b9a\u4ef7\u3001\u6838\u5fc3\u5356\u70b9\u3001\u76ee\u6807\u7528\u6237）\\\\n"<\/span>
\n "- \u5e02\u573a\u89c4\u6a21\u4f30\u7b97\u548c\u589e\u901f\\\\n"<\/span>
\n "- \u7528\u6237\u9700\u6c42 Top5"<\/span>
\n )<\/span>,<\/span>
\n agent=<\/span>agent,<\/span>
\n output_file=<\/span>"outputs\/research.md"<\/span>
\n )<\/span><\/p>\n

def<\/span> create_analysis_task<\/span>(<\/span>agent,<\/span> product_name)<\/span>:<\/span>
\n """\u7b2c\u4e8c\u6b65：\u5206\u6790\u5bf9\u6bd4"""<\/span>
\n return<\/span> Task(<\/span>
\n description=<\/span>f"""
\n \u57fa\u4e8e\u4e0a\u4e00\u6b65\u7684\u8c03\u7814\u6570\u636e，\u5bf9"<\/span>{<\/span>product_name}<\/span><\/span>"\u5e02\u573a\u505a\u6df1\u5ea6\u5206\u6790：
\n 1. \u7ade\u54c1\u5b9a\u4ef7\u7b56\u7565\u5bf9\u6bd4（\u529f\u80fdvs\u4ef7\u683c\u77e9\u9635）
\n 2. \u5404\u7ade\u54c1\u7684\u5dee\u5f02\u5316\u4f18\u52bf\u548c\u5f31\u70b9
\n 3. \u6211\u4eec\u7684\u673a\u4f1a\u7a97\u53e3\u548c\u6f5c\u5728\u5a01\u80c1
\n """<\/span><\/span>,<\/span>
\n expected_output=<\/span>(<\/span>
\n "\u4e00\u4efd\u6570\u636e\u5206\u6790\u62a5\u544a，\u5305\u542b：\\\\n"<\/span>
\n "- \u4ef7\u683c-\u529f\u80fd\u5bf9\u6bd4\u8868\u683c\\\\n"<\/span>
\n "- SWOT\u5206\u6790（\u4f18\u52bf\/\u52a3\u52bf\/\u673a\u4f1a\/\u5a01\u80c1）\\\\n"<\/span>
\n "- 3\u4e2a\u6708\u5185\u7684\u4f18\u5148\u884c\u52a8\u5efa\u8bae"<\/span>
\n )<\/span>,<\/span>
\n agent=<\/span>agent,<\/span>
\n context=<\/span>[<\/span>]<\/span>,<\/span> # \u5c06\u7531 Crew \u81ea\u52a8\u6ce8\u5165\u524d\u5e8f\u4efb\u52a1\u7684\u8f93\u51fa<\/span>
\n output_file=<\/span>"outputs\/analysis.md"<\/span>
\n )<\/span><\/p>\n

def<\/span> create_strategy_task<\/span>(<\/span>agent,<\/span> product_name)<\/span>:<\/span>
\n """\u7b2c\u4e09\u6b65：\u7b56\u7565\u5efa\u8bae"""<\/span>
\n return<\/span> Task(<\/span>
\n description=<\/span>f"""
\n \u57fa\u4e8e\u8c03\u7814\u548c\u5206\u6790\u7ed3\u679c，\u4e3a"<\/span>{<\/span>product_name}<\/span><\/span>"\u4ea7\u54c1\u5236\u5b9a：
\n 1. \u5dee\u5f02\u5316\u5b9a\u4f4d\u5efa\u8bae
\n 2. \u5b9a\u4ef7\u533a\u95f4\u548c\u7406\u7531
\n 3. \u5e02\u573a\u8fdb\u5165\u8def\u5f84（\u7b2c\u4e00\u6b65\u505a\u4ec0\u4e48\u3001\u7b2c\u4e8c\u6b65\u505a\u4ec0\u4e48）
\n 4. \u98ce\u9669\u6e05\u5355\u548c\u5e94\u5bf9\u9884\u6848
\n """<\/span><\/span>,<\/span>
\n expected_output=<\/span>(<\/span>
\n "\u4e00\u4efd\u53ef\u6267\u884c\u7684\u7b56\u7565\u6587\u6863，\u5305\u542b\u5b9a\u4f4d\u3001\u5b9a\u4ef7\u3001\u8def\u5f84\u548c\u98ce\u9669\u56db\u90e8\u5206，"<\/span>
\n "\u6bcf\u90e8\u5206\u6709\u660e\u786e\u7684\u7406\u7531\u548c\u6570\u636e\u652f\u6491\u3002"<\/span>
\n )<\/span>,<\/span>
\n agent=<\/span>agent,<\/span>
\n context=<\/span>[<\/span>]<\/span>,<\/span>
\n output_file=<\/span>"outputs\/strategy.md"<\/span>
\n )<\/span><\/p>\n

\u6b65\u9aa45：\u7ec4\u88c5 Crew<\/h5>\n

\u76ee\u6807：\u628a Agent \u548c Task \u7ec4\u88c5\u6210\u53ef\u6267\u884c\u7684 Crew\u3002<\/p>\n

# crew.py<\/span>
\nfrom<\/span> crewai import<\/span> Crew,<\/span> Process
\nfrom<\/span> crewai import<\/span> LLM
\nfrom<\/span> agents import<\/span> create_researcher,<\/span> create_analyst,<\/span> create_strategist
\nfrom<\/span> tasks import<\/span> create_research_task,<\/span> create_analysis_task,<\/span> create_strategy_task<\/p>\n

def<\/span> create_market_research_crew<\/span>(<\/span>product_name)<\/span>:<\/span>
\n # 1. \u521d\u59cb\u5316 LLM<\/span>
\n llm =<\/span> LLM(<\/span>
\n model=<\/span>"gpt-4o-mini"<\/span>,<\/span> # \u8c03\u8bd5\u7528\u4f4e\u6210\u672c\u6a21\u578b<\/span>
\n temperature=<\/span>0.7<\/span>
\n )<\/span><\/p>\n

# 2. \u521b\u5efa Agent<\/span>
\n researcher =<\/span> create_researcher(<\/span>llm)<\/span>
\n analyst =<\/span> create_analyst(<\/span>llm)<\/span>
\n strategist =<\/span> create_strategist(<\/span>llm)<\/span><\/p>\n

# 3. \u521b\u5efa Task<\/span>
\n research_task =<\/span> create_research_task(<\/span>researcher,<\/span> product_name)<\/span>
\n analysis_task =<\/span> create_analysis_task(<\/span>analyst,<\/span> product_name)<\/span>
\n strategy_task =<\/span> create_strategy_task(<\/span>strategist,<\/span> product_name)<\/span><\/p>\n

# 4. \u7ec4\u88c5 Crew<\/span>
\n return<\/span> Crew(<\/span>
\n agents=<\/span>[<\/span>researcher,<\/span> analyst,<\/span> strategist]<\/span>,<\/span>
\n tasks=<\/span>[<\/span>research_task,<\/span> analysis_task,<\/span> strategy_task]<\/span>,<\/span>
\n process=<\/span>Process.<\/span>sequential,<\/span> # \u6309\u987a\u5e8f\u6267\u884c<\/span>
\n verbose=<\/span>True<\/span>
\n )<\/span><\/p>\n

# main.py (\u6700\u7ec8\u7248\u672c)<\/span>
\nfrom<\/span> dotenv import<\/span> load_dotenv
\nload_dotenv(<\/span>)<\/span><\/p>\n

from<\/span> crew import<\/span> create_market_research_crew<\/p>\n

if<\/span> __name__ ==<\/span> "__main__"<\/span>:<\/span>
\n crew =<\/span> create_market_research_crew(<\/span>"AI\u4ee3\u7801\u52a9\u624b"<\/span>)<\/span>
\n result =<\/span> crew.<\/span>kickoff(<\/span>)<\/span>
\n print<\/span>(<\/span>"\\\\n"<\/span> +<\/span> "="<\/span>*<\/span>60<\/span>)<\/span>
\n print<\/span>(<\/span>"\u6700\u7ec8\u8f93\u51fa："<\/span>)<\/span>
\n print<\/span>(<\/span>"="<\/span>*<\/span>60<\/span>)<\/span>
\n print<\/span>(<\/span>result.<\/span>raw)<\/span><\/p>\n

3.4 \u8fd0\u884c\u7ed3\u679c<\/h4>\n

\u8fd0\u884c python main.py \u540e，\u5178\u578b\u8f93\u51fa\u5982\u4e0b：<\/p>\n

[2026-06-04 10:30:00][DEBUG]: == Working Agent: \u5e02\u573a\u4fe1\u606f\u68c0\u7d22\u4e13\u5bb6
\n[2026-06-04 10:30:00][INFO]: == Starting Task: \u8bf7\u641c\u7d22"AI\u4ee3\u7801\u52a9\u624b"\u76f8\u5173\u7684\u5e02\u573a\u4fe1\u606f…
\n[2026-06-04 10:32:15][DEBUG]: == Agent: \u5e02\u573a\u4fe1\u606f\u68c0\u7d22\u4e13\u5bb6 – Tool calling: None – Reasoning: None
\n[2026-06-04 10:32:45][INFO]: == Task completed successfully.<\/p>\n

[2026-06-04 10:32:45][DEBUG]: == Working Agent: \u6570\u636e\u5206\u6790\u5e08
\n[2026-06-04 10:32:45][INFO]: == Starting Task: \u57fa\u4e8e\u4e0a\u4e00\u6b65\u7684\u8c03\u7814\u6570\u636e，\u5bf9"AI\u4ee3\u7801\u52a9\u624b"\u5e02\u573a\u505a\u6df1\u5ea6\u5206\u6790…
\n[2026-06-04 10:34:30][INFO]: == Task completed successfully.<\/p>\n

[2026-06-04 10:34:30][DEBUG]: == Working Agent: \u7b56\u7565\u5206\u6790\u5e08
\n[2026-06-04 10:34:30][INFO]: == Starting Task: \u57fa\u4e8e\u8c03\u7814\u548c\u5206\u6790\u7ed3\u679c…
\n[2026-06-04 10:36:00][INFO]: == Task completed successfully.<\/p>\n

============================================================
\n\u6700\u7ec8\u8f93\u51fa：
\n============================================================
\n# AI\u4ee3\u7801\u52a9\u624b\u5e02\u573a\u7b56\u7565\u62a5\u544a
\n…<\/p>\n

\u8f93\u51fa\u6587\u4ef6 outputs\/ \u76ee\u5f55\u4e0b\u5c06\u751f\u6210\u4e09\u4efd\u62a5\u544a：research.md\u3001analysis.md\u3001strategy.md\u3002<\/p>\n

3.5 \u6d4b\u8bd5\u9a8c\u8bc1<\/h4>\n

\u4f7f\u7528 pytest \u9a8c\u8bc1\u6838\u5fc3\u7ec4\u4ef6：<\/p>\n

# test_crew.py<\/span>
\nimport<\/span> pytest
\nfrom<\/span> agents import<\/span> create_researcher,<\/span> create_analyst,<\/span> create_strategist
\nfrom<\/span> crewai import<\/span> LLM
\nfrom<\/span> unittest.<\/span>mock import<\/span> MagicMock<\/p>\n

def<\/span> test_agent_creation<\/span>(<\/span>)<\/span>:<\/span>
\n """\u6d4b\u8bd5 Agent \u662f\u5426\u80fd\u6b63\u786e\u521b\u5efa"""<\/span>
\n llm =<\/span> LLM(<\/span>model=<\/span>"gpt-4o-mini"<\/span>)<\/span><\/p>\n

researcher =<\/span> create_researcher(<\/span>llm)<\/span>
\n assert<\/span> researcher.<\/span>role ==<\/span> "\u5e02\u573a\u4fe1\u606f\u68c0\u7d22\u4e13\u5bb6"<\/span>
\n assert<\/span> researcher.<\/span>allow_delegation ==<\/span> False<\/span><\/p>\n

analyst =<\/span> create_analyst(<\/span>llm)<\/span>
\n assert<\/span> analyst.<\/span>role ==<\/span> "\u6570\u636e\u5206\u6790\u5e08"<\/span><\/p>\n

strategist =<\/span> create_strategist(<\/span>llm)<\/span>
\n assert<\/span> strategist.<\/span>allow_delegation ==<\/span> True<\/span><\/p>\n

def<\/span> test_task_dependency<\/span>(<\/span>)<\/span>:<\/span>
\n """\u6d4b\u8bd5 Task \u4e0a\u4e0b\u6587\u4f9d\u8d56\u914d\u7f6e"""<\/span>
\n from<\/span> tasks import<\/span> create_analysis_task
\n from<\/span> crewai import<\/span> Agent<\/p>\n

llm =<\/span> LLM(<\/span>model=<\/span>"gpt-4o-mini"<\/span>)<\/span>
\n agent =<\/span> Agent(<\/span>role=<\/span>"test"<\/span>,<\/span> goal=<\/span>"test"<\/span>,<\/span> backstory=<\/span>"test"<\/span>,<\/span> llm=<\/span>llm)<\/span>
\n task =<\/span> create_analysis_task(<\/span>agent,<\/span> "test_product"<\/span>)<\/span><\/p>\n

assert<\/span> task.<\/span>context ==<\/span> [<\/span>]<\/span> # \u5e94\u4e3a\u7a7a\u5217\u8868，\u8868\u793a\u63a5\u6536\u524d\u5e8f\u4efb\u52a1<\/span>
\n assert<\/span> "expected_output"<\/span> in<\/span> task.<\/span>__dict__<\/p>\n

\u8fd0\u884c\u6d4b\u8bd5：<\/p>\n

pytest test_crew.py -v<\/span><\/p>\n

test_crew.py::test_agent_creation PASSED
\ntest_crew.py::test_task_dependency PASSED<\/p>\n

4. \u9879\u76ee\u603b\u7ed3<\/h3>\n

4.1 \u4f18\u70b9\u4e0e\u7f3a\u70b9<\/h4>\n\n\u7ef4\u5ea6CrewAI\u4f20\u7edf\u5355\u4f53 AgentLangChain\/LangGraph<\/tr>\n\n\n\n\n\n\n\n
\u4e0a\u624b\u96be\u5ea6<\/td>\n\u4f4e，\u89d2\u8272-\u4efb\u52a1-\u5de5\u5177\u4e09\u8981\u7d20\u6e05\u6670<\/td>\n\u4f4e，\u4f46\u6269\u5c55\u6027\u5dee<\/td>\n\u4e2d\u9ad8，\u9700\u7406\u89e3\u94fe\u548c\u56fe<\/td>\n<\/tr>\n
\u591a\u89d2\u8272\u534f\u4f5c<\/td>\n\u539f\u751f\u652f\u6301，Process \u81ea\u52a8\u7f16\u6392<\/td>\n\u9700\u624b\u52a8\u7f16\u7801\u534f\u4f5c\u903b\u8f91<\/td>\n\u9700\u624b\u52a8\u5b9a\u4e49\u8282\u70b9\u548c\u8fb9<\/td>\n<\/tr>\n
\u5de5\u5177\u96c6\u6210<\/td>\n\u5185\u7f6e\u5de5\u5177 + \u81ea\u5b9a\u4e49\u7b80\u5355<\/td>\n\u9700\u81ea\u884c\u5c01\u88c5<\/td>\n\u5de5\u5177\u94fe\u4e30\u5bcc\u4f46\u590d\u6742<\/td>\n<\/tr>\n
\u8f93\u51fa\u63a7\u5236<\/td>\nexpected_output + Guardrail<\/td>\n\u9760 prompt \u7ea6\u675f<\/td>\n\u9700\u989d\u5916\u89e3\u6790\u5668<\/td>\n<\/tr>\n
\u5927\u578b\u9879\u76ee<\/td>\nFlow \u53ef\u7f16\u6392\u591a\u4e2a Crew<\/td>\n\u96be\u4ee5\u7ba1\u7406<\/td>\nGraph \u9002\u5408\u590d\u6742 DAG<\/td>\n<\/tr>\n
\u6a21\u578b\u4f9d\u8d56\u5ea6\u9ad8\u65f6\u7684\u4e0d\u786e\u5b9a\u6027<\/td>\n\u4ecd\u662f\u6311\u6218<\/td>\n\u540c\u6837<\/td>\n\u540c\u6837<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

4.2 \u9002\u7528\u573a\u666f<\/h4>\n

\u63a8\u8350\u4f7f\u7528 CrewAI \u7684\u573a\u666f：<\/p>\n

  • \u9700\u8981\u591a\u89d2\u8272\u5206\u5de5\u7684\u4e1a\u52a1\u6d41\u7a0b（\u8c03\u7814\u2192\u5206\u6790\u2192\u51fa\u62a5\u544a）<\/li>\n
  • AI \u8f85\u52a9\u7684\u5de5\u5355\u5206\u62e3\u548c\u8def\u7531<\/li>\n
  • \u591a\u6e90\u4fe1\u606f\u805a\u5408\u548c\u6458\u8981\u751f\u6210<\/li>\n
  • \u9700\u8981\u7ed3\u5408\u79c1\u6709\u77e5\u8bc6\u5e93\u7684\u95ee\u7b54\u7cfb\u7edf<\/li>\n
  • \u9700\u4eba\u5de5\u786e\u8ba4\u5173\u952e\u51b3\u7b56\u8282\u70b9\u7684\u81ea\u52a8\u5316\u6d41\u7a0b<\/li>\n

    \u4e0d\u63a8\u8350\u4f7f\u7528\u7684\u573a\u666f：<\/p>\n

  • \u7eaf\u786e\u5b9a\u6027\u8ba1\u7b97（\u5982\u5de5\u8d44\u6838\u7b97）\u2014\u2014\u4f20\u7edf\u4ee3\u7801\u66f4\u53ef\u9760<\/li>\n
  • \u5b9e\u65f6\u6027\u8981\u6c42\u6781\u9ad8\u7684\u573a\u666f（<100ms）\u2014\u2014LLM \u8c03\u7528\u5ef6\u8fdf\u4e0d\u53ef\u63a7<\/li>\n

    4.3 \u6ce8\u610f\u4e8b\u9879<\/h4>\n
  • API Key \u5b89\u5168：\u6c38\u8fdc\u4e0d\u8981\u5c06 API Key \u786c\u7f16\u7801\u5728\u4ee3\u7801\u4e2d，\u4f7f\u7528 .env \u6587\u4ef6\u5e76\u52a0\u5165 .gitignore\u3002<\/li>\n
  • \u6a21\u578b\u9009\u62e9：\u8c03\u8bd5\u9636\u6bb5\u4f7f\u7528 gpt-4o-mini \u6216 gpt-3.5-turbo \u964d\u4f4e\u6210\u672c，\u751f\u4ea7\u73af\u5883\u6309\u4efb\u52a1\u590d\u6742\u5ea6\u9009\u578b\u3002<\/li>\n
  • Verbose \u6a21\u5f0f：\u5f00\u53d1\u9636\u6bb5\u5f00\u542f verbose=True \u4fbf\u4e8e\u6392\u67e5，\u751f\u4ea7\u73af\u5883\u5173\u95ed\u4ee5\u51cf\u5c11\u65e5\u5fd7\u91cf\u3002<\/li>\n
  • allow_delegation：\u591a\u6570 Agent \u5e94\u8bbe\u4e3a False，\u4ec5 Manager \u89d2\u8272\u5f00\u542f，\u907f\u514d\u610f\u5916\u59d4\u6258\u5bfc\u81f4 Token \u6d6a\u8d39\u3002<\/li>\n

    4.4 \u5e38\u89c1\u8e29\u5751\u7ecf\u9a8c<\/h4>\n

    \u6848\u4f8b1：Agent \u89d2\u8272\u63cf\u8ff0\u592a\u6cdb\u5bfc\u81f4\u8f93\u51fa\u7a7a\u6d1e<\/p>\n

    \u73b0\u8c61：Agent \u7684 role \u5199\u6210"AI \u52a9\u624b"，\u8f93\u51fa\u603b\u662f\u4e00\u5806\u5e9f\u8bdd\u3002 \u6839\u56e0：role\/goal\/backstory \u8d8a\u5177\u4f53，\u6a21\u578b\u8d8a\u6709\u65b9\u5411\u3002\u89d2\u8272\u63cf\u8ff0\u5b9e\u9645\u4e0a\u662f\u5728\u4e3a\u6a21\u578b\u7684 System Prompt \u63d0\u4f9b\u4e0a\u4e0b\u6587\u7ea6\u675f\u3002 \u89e3\u51b3：\u5c06 role \u6539\u4e3a"\u7ade\u54c1\u5b9a\u4ef7\u5206\u6790\u5e08"，goal \u6539\u4e3a"\u5bf9\u6bd4 3 \u6b3e\u7ade\u54c1\u7684\u5b9a\u4ef7\u7b56\u7565\u5e76\u7ed9\u51fa\u5b9a\u4ef7\u533a\u95f4\u5efa\u8bae"\u3002<\/p>\n

    \u6848\u4f8b2：Task context \u4e0d\u4f20\u5bfc\u81f4\u4fe1\u606f\u65ad\u88c2<\/p>\n

    \u73b0\u8c61：\u7b2c\u4e8c\u4e2a Task \u7684\u8f93\u51fa\u5b8c\u5168\u6ca1\u6709\u53c2\u8003\u7b2c\u4e00\u4e2a Task \u7684\u7ed3\u679c\u3002 \u6839\u56e0：Task \u7684 context \u53c2\u6570\u6ca1\u6709\u8bbe\u7f6e\u4e3a\u63a5\u6536\u524d\u5e8f\u4efb\u52a1\u8f93\u51fa\u3002 \u89e3\u51b3：\u5c06 context=[] \u914d\u7f6e\u597d，Crew \u4f1a\u81ea\u52a8\u628a\u524d\u5e8f Task \u7684 output \u6ce8\u5165\u5f53\u524d Task \u7684\u4e0a\u4e0b\u6587\u3002<\/p>\n

    \u6848\u4f8b3：Process.sequential \u6267\u884c"\u5361\u4f4f"<\/p>\n

    \u73b0\u8c61：Task \u6267\u884c\u5230\u4e00\u534a\u957f\u65f6\u95f4\u65e0\u54cd\u5e94\u3002 \u6839\u56e0：LLM API \u8d85\u65f6\u6216\u901f\u7387\u9650\u5236\u3002 \u89e3\u51b3：\u5728 LLM \u521d\u59cb\u5316\u65f6\u8bbe\u7f6e timeout=120 \u548c max_retries=3；\u68c0\u67e5 API \u8d26\u6237\u4f59\u989d\u548c\u901f\u7387\u9650\u5236\u3002<\/p>\n

    4.5 \u601d\u8003\u9898<\/h4>\n
  • \n

    \u5982\u679c\u8ba9\u4e00\u4e2a Agent \u540c\u65f6\u627f\u62c5"\u4fe1\u606f\u68c0\u7d22"\u548c"\u6570\u636e\u5206\u6790"\u4e24\u4e2a\u89d2\u8272，\u4f1a\u6709\u4ec0\u4e48\u95ee\u9898？\u4ec0\u4e48\u60c5\u51b5\u4e0b\u8fd9\u79cd\u8bbe\u8ba1\u53cd\u800c\u662f\u66f4\u597d\u7684\u9009\u62e9？（\u63d0\u793a：\u8003\u8651 Token \u6210\u672c\u3001\u4e0a\u4e0b\u6587\u957f\u5ea6\u9650\u5236\u3001\u89d2\u8272\u4e13\u6ce8\u5ea6）<\/p>\n<\/li>\n

  • \n

    Sequential Process \u548c Hierarchical Process \u7684\u6838\u5fc3\u5dee\u5f02\u662f\u4ec0\u4e48？\u5728\u4ec0\u4e48\u4e1a\u52a1\u573a\u666f\u4e0b\u4f60\u4f1a\u9009\u62e9 Hierarchical？（\u63d0\u793a：\u8003\u8651\u4efb\u52a1\u7684\u4e0d\u786e\u5b9a\u6027\u3001Agent \u4e4b\u95f4\u7684\u4f9d\u8d56\u5173\u7cfb\u3001\u6210\u672c\u7684\u5dee\u5f02）<\/p>\n<\/li>\n

    \u7b54\u6848\u5c06\u5728\u7b2c2\u7ae0\u53ca\u540e\u7eed\u7ae0\u8282\u4e2d\u9010\u6b65\u63ed\u6653\u3002<\/p>\n

    \u5ef6\u4f38\u9605\u8bfb\u4e0e\u8d44\u6e90<\/h3>\n

    10\u500d\u5f00\u53d1\u8005\u7684 Dify \u9b54\u6cd5\u4e66：\u4ece\u96f6\u6784\u5efa\u5168\u6808 AI \u5e94\u7528 \u540e\u7aef\u5de5\u7a0b\u5e08\u8f6c\u578bAI\u7b2c\u4e00\u8bfe-Ollama \u4e0e\u79c1\u6709\u5316\u5927\u6a21\u578b\u5b9e\u6218 \u5927\u578b\u8bed\u8a00\u6a21\u578b(LLM) vLLM \u9ad8\u6027\u80fd\u63a8\u7406\u843d\u5730\u5b9e\u6218 Agent\u5f00\u53d1\u4e4bLlamaIndex \u5b9e\u6218\u4fee\u70bc\u4e0e\u6e90\u7801\u8fdb\u9636 \u5927\u8bed\u8a00\u6a21\u578bTransformers \u5b9e\u6218\u4fee\u70bc\u4e0e\u6e90\u7801\u5256\u6790<\/p>\n","protected":false},"excerpt":{"rendered":"

    1. \u9879\u76ee\u80cc\u666f
    \n\u5982\u679c\u4f60\u662f\u4e00\u5bb6\u4e2d\u578b\u7535\u5546\u516c\u53f8\u7684\u6280\u672f\u8d1f\u8d23\u4eba，\u4f60\u4f1a\u53d1\u73b0\u56e2\u961f\u6b63\u5728\u7ecf\u5386\u4e00\u573a\\”AI \u7126\u8651\\”\u3002\u8fd0\u8425\u90e8\u95e8\u5e0c\u671b\u7528\u5927\u6a21\u578b\u81ea\u52a8\u751f\u6210\u5546\u54c1\u6587\u6848，\u5ba2\u670d\u56e2\u961f\u60f3\u8981\u4e00\u4e2a\u80fd\u56de\u7b54 80% \u5e38\u89c1\u95ee\u9898\u7684\u673a\u5668\u4eba，\u6570\u636e\u5206\u6790\u5e08\u5219\u60f3\u628a\u6bcf\u5468\u4e94\u7684\u62a5\u8868\u4ece\\”\u624b\u52a8\u62d6\u6570\u2014\u2014Excel \u753b\u56fe\u2014\u2014\u5199\u7ed3\u8bba\\”\u538b\u7f29\u5230\u4e00\u952e\u751f\u6210\u3002
    \n\u4e8e\u662f\u4f60\u8ba9\u56e2\u961f\u5404\u81ea\u5c1d\u8bd5：\u8fd0\u8425\u7528 GPT-4 \u5199 prompt \u751f\u6210\u6587\u6848，\u7ed3\u679c\u8d28\u91cf\u5ffd\u9ad8\u5ffd\u4f4e\u2014\u2014\u540c\u6837\u7684 prompt，\u6709\u65f6\u8f93\u51fa<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[13094,13095,74,5095,5079],"topic":[],"class_list":["post-115171","post","type-post","status-publish","format-standard","hentry","category-server","tag-crew","tag-process","tag-agent","tag-crewai","tag-task"],"yoast_head":"\n\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/115171.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"1. \u9879\u76ee\u80cc\u666f \u5982\u679c\u4f60\u662f\u4e00\u5bb6\u4e2d\u578b\u7535\u5546\u516c\u53f8\u7684\u6280\u672f\u8d1f\u8d23\u4eba，\u4f60\u4f1a\u53d1\u73b0\u56e2\u961f\u6b63\u5728\u7ecf\u5386\u4e00\u573a"AI \u7126\u8651"\u3002\u8fd0\u8425\u90e8\u95e8\u5e0c\u671b\u7528\u5927\u6a21\u578b\u81ea\u52a8\u751f\u6210\u5546\u54c1\u6587\u6848，\u5ba2\u670d\u56e2\u961f\u60f3\u8981\u4e00\u4e2a\u80fd\u56de\u7b54 80% \u5e38\u89c1\u95ee\u9898\u7684\u673a\u5668\u4eba，\u6570\u636e\u5206\u6790\u5e08\u5219\u60f3\u628a\u6bcf\u5468\u4e94\u7684\u62a5\u8868\u4ece"\u624b\u52a8\u62d6\u6570\u2014\u2014Excel \u753b\u56fe\u2014\u2014\u5199\u7ed3\u8bba"\u538b\u7f29\u5230\u4e00\u952e\u751f\u6210\u3002 \u4e8e\u662f\u4f60\u8ba9\u56e2\u961f\u5404\u81ea\u5c1d\u8bd5：\u8fd0\u8425\u7528 GPT-4 \u5199 prompt \u751f\u6210\u6587\u6848，\u7ed3\u679c\u8d28\u91cf\u5ffd\u9ad8\u5ffd\u4f4e\u2014\u2014\u540c\u6837\u7684 prompt，\u6709\u65f6\u8f93\u51fa\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/115171.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2026-10-11T04:08:55+00:00\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/115171.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/115171.html\",\"name\":\"\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2026-10-11T04:08:55+00:00\",\"dateModified\":\"2026-10-11T04:08:55+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/115171.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/115171.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/115171.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/115171.html","og_locale":"zh_CN","og_type":"article","og_title":"\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"1. \u9879\u76ee\u80cc\u666f \u5982\u679c\u4f60\u662f\u4e00\u5bb6\u4e2d\u578b\u7535\u5546\u516c\u53f8\u7684\u6280\u672f\u8d1f\u8d23\u4eba，\u4f60\u4f1a\u53d1\u73b0\u56e2\u961f\u6b63\u5728\u7ecf\u5386\u4e00\u573a\"AI \u7126\u8651\"\u3002\u8fd0\u8425\u90e8\u95e8\u5e0c\u671b\u7528\u5927\u6a21\u578b\u81ea\u52a8\u751f\u6210\u5546\u54c1\u6587\u6848，\u5ba2\u670d\u56e2\u961f\u60f3\u8981\u4e00\u4e2a\u80fd\u56de\u7b54 80% \u5e38\u89c1\u95ee\u9898\u7684\u673a\u5668\u4eba，\u6570\u636e\u5206\u6790\u5e08\u5219\u60f3\u628a\u6bcf\u5468\u4e94\u7684\u62a5\u8868\u4ece\"\u624b\u52a8\u62d6\u6570\u2014\u2014Excel \u753b\u56fe\u2014\u2014\u5199\u7ed3\u8bba\"\u538b\u7f29\u5230\u4e00\u952e\u751f\u6210\u3002 \u4e8e\u662f\u4f60\u8ba9\u56e2\u961f\u5404\u81ea\u5c1d\u8bd5：\u8fd0\u8425\u7528 GPT-4 \u5199 prompt \u751f\u6210\u6587\u6848，\u7ed3\u679c\u8d28\u91cf\u5ffd\u9ad8\u5ffd\u4f4e\u2014\u2014\u540c\u6837\u7684 prompt，\u6709\u65f6\u8f93\u51fa","og_url":"https:\/\/www.wsisp.com\/helps\/115171.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2026-10-11T04:08:55+00:00","author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"6 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/115171.html","url":"https:\/\/www.wsisp.com\/helps\/115171.html","name":"\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2026-10-11T04:08:55+00:00","dateModified":"2026-10-11T04:08:55+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/115171.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/115171.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/115171.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"\u7b2c1\u7ae0\uff1aCrewAI \u672f\u8bed\u5168\u666f\u4e0e\u591a\u667a\u80fd\u4f53\u5de5\u4f5c\u539f\u7406"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/115171","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=115171"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/115171\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=115171"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=115171"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=115171"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=115171"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}