[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"product:open-viking":3},{"id":4,"name":5,"slug":6,"slogan":7,"description":8,"organization":9,"images":10,"avatar_url":11,"image_url":12,"website_url":13,"access_links":14,"status":19,"build_by":20,"categories":21,"stats":28,"upload_relationship":32,"create_at":33,"update_at":34,"owner_id":29,"upload_user_id":30,"reviewer_id":30,"video_info":35,"prerelease_cover_url":36,"scheduled_publish_at":35,"user_scheduled":37,"emoji_reacts":38,"tag":35,"pass_review":35,"user_star":37},1465,"OpenViking","open-viking","专为 AI Agent 设计的开源上下文数据库。","{\"type\": \"doc\", \"content\": [{\"type\": \"paragraph\", \"content\": [{\"text\": \"【产品介绍】\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"OpenViking 是一款专为 AI Agent 设计的开源上下文数据库。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"它摒弃传统 RAG 的碎片化向量存储模式，创新性地采用“文件系统范式”统一管理 Agent 的记忆、资源与技能。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"通过结构化的信息组织，OpenViking 致力于解决上下文碎片化、Token 消耗过大及检索过程不可解释等难题，助力开发者构建具有长久记忆与自我演进能力的智能 Agent。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\"}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"【产品功能】\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"文件系统范式：通过 viking：\u002F\u002F 协议将上下文映射为虚拟文件目录，支持对记忆和资源进行标准化操作与统一管理。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"分层上下文加载：采用 L0（摘要）\u002FL1（概览）\u002FL2（详情）三级加载策略，按需获取信息，大幅降低 Token 成本。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"目录递归检索：结合意图分析、向量定位与递归下钻，提供比单一向量检索更精准、更具全局观的结构化检索。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"可观测与自演进：支持检索轨迹的全程可视化追踪，并能从会话交互中自动提取经验，实现 Agent 记忆的自我迭代。\", \"type\": \"text\"}]}, {\"type\": \"paragraph\"}, {\"type\": \"paragraph\", \"content\": [{\"text\": \"项目开源地址：\", \"type\": \"text\"}, {\"text\": \"https:\u002F\u002Fgithub.com\u002Fvolcengine\u002FOpenViking\", \"type\": \"text\", \"marks\": [{\"type\": \"link\", \"attrs\": {\"rel\": \"noopener noreferrer nofollow\", \"href\": \"https:\u002F\u002Fgithub.com\u002Fvolcengine\u002FOpenViking\", \"class\": null, \"target\": \"_blank\"}}]}]}]}","字节跳动","https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1771480710_b852cfd7-2400-4ea4-a393-86d3edb7a417.png;https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1771480710_c4e5cd45-24b2-42dd-ba3a-bdb4056b6c9e.png;https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1771480710_c8b9c12e-d051-497d-89aa-1c30b03d5d4e.png","https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1771480702_7927afc8-51dd-4667-822d-0a56d42b77dc.png","https:\u002F\u002Fwatcha.tos-cn-beijing.volces.com\u002Fprod\u002Fuser\u002Fuploads\u002F2_1771480706_16025d8c-cea2-4166-9966-e3464198a8a8.png","https:\u002F\u002Fopenviking.ai\u002F",{"items":15},[16],{"platform":17,"url":13,"is_primary":18},"WEB",true,"PUBLISHED",[],[22,25],{"id":23,"name":24},6,"编程开发",{"id":26,"name":27},14,"效率工具",{"upvotes":29,"stars":30,"review_count":29,"reply_count":29,"update_at":31},0,2,"2026-08-24T07:09:26.902Z","THIRD_PARTY","2026-02-19T05:58:36.464Z","2026-02-19T05:59:02.244Z",null,"",false,{"reacts":39},[]]