[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-82780":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":8,"htmlUrl":8,"language":9,"languages":8,"totalLinesOfCode":8,"stars":10,"forks":11,"watchers":12,"openIssues":13,"contributorsCount":14,"subscribersCount":14,"size":14,"stars1d":15,"stars7d":16,"stars30d":17,"stars90d":14,"forks30d":14,"starsTrendScore":18,"compositeScore":19,"rankGlobal":8,"rankLanguage":8,"license":8,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":22,"hasPages":20,"topics":23,"createdAt":8,"pushedAt":8,"updatedAt":24,"readmeContent":25,"aiSummary":26,"trendingCount":14,"starSnapshotCount":14,"syncStatus":15,"lastSyncTime":27,"discoverSource":28},82780,"chokepoint-atlas","qiuqiubuchongle-cloud\u002Fchokepoint-atlas","qiuqiubuchongle-cloud",null,"Python",625,130,109,1,0,2,77,516,24,10.35,false,"main",true,[],"2026-06-12 02:04:27","# Chokepoint Atlas\n\n`Chokepoint Atlas` is a reusable research skill for finding real bottlenecks in AI infrastructure supply chains.\n\n## 中文\n\n**Chokepoint Atlas** 是一个面向 AI 基础设施研究的 Agent Skill。\n\n它不是上来就给股票代码的选股 prompt。  \n它更像一个研究助手，专门帮你做一件事：\n\n**把热门 AI 叙事拆成真实供应链，再找出最容易卡脖子的那一层。**\n\n核心流程很简单：\n\n1. 先确认终端系统\n2. 再拆供应链栈\n3. 找真正的瓶颈\n4. 用财报、研报、产业新闻交叉验证\n5. 先给方向，再给名字\n\n它适合拿来研究：\n\n- AI 光通信\n- 封装与测试\n- 数据中心供电\n- 液冷与热管理\n- 机器人供应链\n- 其他 AI 基础设施方向\n\n它不适合：\n\n- 直接要短线代码\n- 只想看情绪和热度\n- 不想看逻辑、只想抄答案\n\n一句话说：\n\n**Chokepoint Atlas 不是帮你追热点，而是帮你先找到真正会堵车的地方。**\n\n## English\n\n**Chokepoint Atlas** is an agent skill for AI infrastructure research.\n\nIt is not a stock-picking prompt that jumps straight to tickers.  \nIt is a research workflow designed to do one thing well:\n\n**turn broad AI narratives into real supply-chain maps, then identify the layer most likely to become a bottleneck.**\n\nThe workflow is straightforward:\n\n1. Define the end system\n2. Map the supply-chain stack\n3. Find the real constraint\n4. Cross-check with earnings, reports, and industry news\n5. Output the direction first, then candidate names\n\nIt is useful for researching:\n\n- AI optical interconnect\n- packaging and testing\n- datacenter power delivery\n- liquid cooling and thermal management\n- humanoid robotics supply chains\n- other AI infrastructure lanes\n\nIt is not meant for:\n\n- instant ticker dumping\n- momentum-only workflows\n- users who want answers without thesis building\n\nIn one line:\n\n**Chokepoint Atlas does not help you chase noise. It helps you find where the system will actually break first.**\n\n## Install\n\nInstall the skill as `ai-supply-chain-bottleneck-hunter`:\n\n```bash\npython3 ~\u002F.codex\u002Fskills\u002F.system\u002Fskill-installer\u002Fscripts\u002Finstall-skill-from-github.py \\\n  --repo qiuqiubuchongle-cloud\u002Fchokepoint-atlas \\\n  --path . \\\n  --name ai-supply-chain-bottleneck-hunter\n```\n\nIf the default download mode fails, use git mode:\n\n```bash\npython3 ~\u002F.codex\u002Fskills\u002F.system\u002Fskill-installer\u002Fscripts\u002Finstall-skill-from-github.py \\\n  --repo qiuqiubuchongle-cloud\u002Fchokepoint-atlas \\\n  --path . \\\n  --name ai-supply-chain-bottleneck-hunter \\\n  --method git\n```\n\nThen restart Codex so the new skill is picked up.\n\n## Repository Contents\n\n- [SKILL.md](.\u002FSKILL.md): the main skill\n- [中文产品说明](.\u002Fdocs\u002FPRODUCT_CN.md)\n- [English product description](.\u002Fdocs\u002FPRODUCT_EN.md)\n- [2.0 产品规划](.\u002Fdocs\u002FPRODUCT_PLAN_V2_CN.md)\n- [Product manual](.\u002Freferences\u002Fproduct-manual.md)\n- [Infographic copy](.\u002Fdocs\u002FINFOGRAPHIC_COPY.md)\n\n## Local MVP\n\nThis repo now includes a small local MVP that turns structured lane input into a reusable research pack.\n\nRun:\n\n```bash\npython3 scripts\u002Fbuild_research_pack.py \\\n  --input examples\u002Fai_factory_lane_input.json \\\n  --output out\u002Fai_factory_demo\n```\n\nGenerated outputs include:\n\n- `research_pack.json`\n- `quick_scan.md`\n- `evidence_memo.md`\n- `graph.json`\n- `graph.mmd`\n- `graph_mermaid.md`\n- `graph_card.md`\n- `scorecard.json`\n- `validation_report.json`\n- `catalyst_watch.md`\n\n## Lane Comparison\n\nYou can also compare multiple lanes in one run:\n\n```bash\npython3 scripts\u002Fcompare_lanes.py \\\n  --input examples\u002Flane_compare_input.json \\\n  --output out\u002Flane_compare_demo\n```\n\nGenerated outputs include:\n\n- `lane_ranking.json`\n- `lane_details.json`\n- `ranked_lane_table.md`\n- `lane_compare_memo.md`\n\n## Source Pipeline\n\nYou can also start from a looser source bundle and let the repo build the draft input plus the final pack:\n\n```bash\npython3 scripts\u002Frun_source_pipeline.py \\\n  --input examples\u002Fsource_bundle_input.json \\\n  --output out\u002Fsource_pipeline_demo\n```\n\nThis pipeline produces:\n\n- `01_draft\u002Fdraft_pack_input.json`\n- `01_draft\u002Fextraction_report.json`\n- `02_final_pack\u002Fresearch_pack.json`\n- `02_final_pack\u002Fquick_scan.md`\n- `02_final_pack\u002Fevidence_memo.md`\n- `02_final_pack\u002Fgraph.json`\n- `02_final_pack\u002Fgraph.mmd`\n- `02_final_pack\u002Fgraph_mermaid.md`\n- `02_final_pack\u002Fscorecard.json`\n- `02_final_pack\u002Fvalidation_report.json`\n\n## Product Snapshot\n\n![AI Bottleneck Hunter infographic](.\u002Fassets\u002Fai-bottleneck-hunter-infographic.png)\n","Chokepoint Atlas 是一个面向 AI 基础设施研究的工具，旨在将广泛的 AI 叙事拆解为真实的供应链，并识别出最可能成为瓶颈的环节。其核心功能包括确认终端系统、拆解供应链栈、找到真正瓶颈，并通过财报、研报和产业新闻进行交叉验证，最终提供方向和具体名称。适用于研究 AI 光通信、封装与测试、数据中心供电、液冷与热管理、机器人供应链等 AI 基础设施领域。它不适合用于直接获取短线股票代码或仅关注市场情绪的用户。简而言之，Chokepoint Atlas 专注于帮助用户找到系统中真正会“卡脖子”的地方，而不是追逐市场热点。","2026-06-11 04:09:13","CREATED_QUERY"]