Open data开放数据
Open data开放数据
Every dataset that powers LynxFlow is downloadable as JSON under /data/ — no auth, no rate limit, no paywall. Build dashboards, feed agents, mirror the directory. License: CC-BY 4.0; please attribute www.lynxflow.co.
LynxFlow 的全部数据集都以 JSON 挂在 /data/ 下,免认证、无限流、不付费。做仪表盘、喂给 Agent、镜像目录都行。许可:CC-BY 4.0,请注明来源 www.lynxflow.co。
/data/tools.json~191KBon edits (~weekly)随编辑(约每周)The full directory: 389 AI tools across 19 categories — bilingual taglines, tags, pricing, cn_ok (China direct access) and a verified url per tool. The canonical dataset the site renders from.完整目录:19 类 389 个 AI 工具——双语标语、标签、定价、国内直连标记(cn_ok)与已探活链接。站点即由此数据渲染。
Schema sketch结构示意
{ categories: [{ category_key, name_en, name_cn, icon, tools: [{ slug, name, url, tagline_en, tagline_cn, pricing, cn_ok, tags[], pick }] }] }/data/health.json~76KBdaily 02:00 (cron)每日 02:00(cron)Liveness probes for every tool url: status (up/blocked/down/unverified), http_status, response_time_ms, uptime_percent, risk (active/risk/inactive), and a 30-entry history. GFW-honest — probed through a global-egress proxy, so geo-blocked ≠ dead.每个工具链接的探活:status(up/blocked/down/unverified)、http_status、响应时间、uptime_percent、risk(active/risk/inactive)与 30 条历史。GFW 诚实——经全球出口代理探活,地理屏蔽≠死链。
Schema sketch结构示意
{ generated_at, n_tools, n_up, n_blocked, n_down, n_inactive, probes: { slug: { status, http_status, response_time_ms, last_check, last_ok, uptime_percent, risk, history[] } } }/data/scenarios.json~15KBon edits随编辑17 "job-to-be-done" scenarios (make-slides, write-copy, make-image…) each with bilingual intro and match{categories,tags_any,slugs} rules that resolve to ranked tools at build time — powers /s/ pages.17 个"按任务找工具"场景(做 PPT、写文案、画图…)各带双语导语与 match{categories,tags_any,slugs} 规则,构建时解析为排好序的工具——驱动 /s/ 页。
Schema sketch结构示意
{ scenarios: [{ slug, name_en, name_cn, query_en, query_cn, desc_en, desc_cn, intro_en, intro_cn, match:{categories,tags_any,slugs}, priority[] }] }/data/vs.json~7KBon edits随编辑17 head-to-head comparison pairs (midjourney-vs-dalle-3, claude-vs-chatgpt…) with bilingual editorial intros. Powers /vs/ pages; tool slugs resolve to full data in tools.json.17 组并排对比(midjourney-vs-dalle-3、claude-vs-chatgpt…)各带双语编辑导语。驱动 /vs/ 页;工具 slug 在 tools.json 里解析为完整数据。
Schema sketch结构示意
{ pairs: [{ a, b, intro_en, intro_cn }] } // url slug = a + "-vs-" + b/data/workflows.json~10KBon edits随编辑6 step-by-step AI workflows (build-saas-mvp, make-presentation, clone-voice…) — each step names the tool slugs that fit and why. Powers /w/ pages.6 条分步 AI 工作流(搭 SaaS、做 PPT、克隆声音…)每步给出最合适的工具 slug 与理由。驱动 /w/ 页。
Schema sketch结构示意
{ workflows: [{ slug, name_en, name_cn, query_en, query_cn, desc_en, desc_cn, intro_en, intro_cn, steps: [{ goal_en, goal_cn, tools[], why_en, why_cn }] }] }↗