[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-93046":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":8,"htmlUrl":8,"language":8,"languages":8,"totalLinesOfCode":8,"stars":9,"forks":10,"watchers":11,"openIssues":10,"contributorsCount":10,"subscribersCount":10,"size":10,"stars1d":10,"stars7d":10,"stars30d":12,"stars90d":10,"forks30d":10,"starsTrendScore":10,"compositeScore":13,"rankGlobal":8,"rankLanguage":8,"license":14,"archived":15,"fork":15,"defaultBranch":16,"hasWiki":17,"hasPages":15,"topics":18,"createdAt":8,"pushedAt":8,"updatedAt":19,"readmeContent":20,"aiSummary":21,"trendingCount":10,"starSnapshotCount":10,"syncStatus":22,"lastSyncTime":23,"discoverSource":24},93046,"bloomy-linkedin-outreach","Litun\u002Fbloomy-linkedin-outreach","Litun",null,116,0,115,1,37.1,"MIT License",false,"main",true,[],"2026-07-22 04:02:08","# bloomy-linkedin-outreach\n\nA template for running a self-driving LinkedIn outreach campaign with an AI agent, using the **Bloomy LinkedIn MCP** to act on your LinkedIn account. Fill in your goal, your offer, and who you're targeting; your agent's scheduled\u002Frecurring tasks handle sourcing, connection requests, intros, follow-ups, and replies - rate-limited, resumable, and observable the whole way.\n\nWorks with any AI agent that supports remote MCP servers and scheduled\u002Frecurring tasks - Claude Cowork is the reference setup. Built by running a real campaign for a week on exactly this structure.\n\n> **AI agent reading this because a user pointed you here?** Fetch this repo and follow `AGENTS.md`.\n\n## Quick start\n\nOpen this project in your AI agent and say **\"let's set up the LinkedIn outreach\"** - it works through `SETUP.md` for you, including checking your Bloomy connection first. `SETUP.md` is the only place setup guidance lives, and it's a one-time thing: day-to-day operation never reads it.\n\n## What's inside\n\n| File | Purpose |\n|---|---|\n| `AGENTS.md` | tells the agent which mode it's in: one-time setup vs. regular runs |\n| `SETUP.md` | the one-time setup guide - run once, not touched again |\n| `CAMPAIGN.md` | your goal, how the system works, rate limits, sequence timing, draft-vs-auto mode |\n| `PRODUCT.md`, `ICP.md`, `TEMPLATES.md` | you fill these in: what you're offering, who you're targeting, how you talk |\n| `outreach_run.md` | the procedure each scheduled outreach run follows |\n| `weekly_review_run.md` | the procedure the weekly funnel-review run follows |\n| `prospects.csv`, `log.md` | campaign state, written by the agent as it runs |\n\n## A note on privacy\n\nOnce configured, this campaign writes real people's LinkedIn data into `prospects.csv` and `log.md`. Don't run it inside a clone that's still connected to this public template's GitHub remote - see SETUP.md step 0.\n\n## License\n\nMIT - see [LICENSE](LICENSE).\n","这是一个用于自动化LinkedIn营销外联的AI代理模板项目，通过集成Bloomy LinkedIn MCP服务实现端到端的自主外联流程。核心功能包括目标设定、ICP筛选、连接请求、个性化开场白、多轮跟进及回复处理，支持速率限制、断点续跑与全流程可观测性。项目采用模块化Markdown配置（如CAMPAIGN.md、TEMPLATES.md），适配支持远程MCP协议和定时任务的AI代理（如Claude Cowork），适用于B2B销售、SaaS获客、招聘拓展等需规模化、合规化LinkedIn触达的场景。",2,"2026-07-11 02:30:44","CREATED_QUERY"]