[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"pack-detail-ai-side-hustle-kit-fr":3,"seo:pack:ai-side-hustle-kit:fr":99},{"code":4,"message":5,"data":6},200,"操作成功",{"pack":7},{"slug":8,"icon":9,"tone":10,"status":11,"status_label":12,"title":13,"description":14,"items":15,"install_cmd":98},"ai-side-hustle-kit","💼","#10B981","new","Nouveau · cette semaine","Kit IA pour Side Hustle","Dix picks pour qu'un fondateur avec un job de jour mette des agents au travail sur un side product — recherche, scraping, contenu, outreach, automatisation, mémoire. Livré dimanche.",[16,28,38,46,54,62,69,76,83,90],{"id":17,"uuid":18,"slug":19,"title":20,"description":21,"author_name":22,"view_count":23,"vote_count":24,"lang_type":25,"type":26,"type_label":27},25,"23330210-b26a-4d97-ad97-1735c203eaa6","gpt-researcher-autonomous-research-report-agent-23330210","GPT Researcher — Autonomous Research Report Agent","AI agent that generates detailed research reports from a single query. Searches multiple sources, synthesizes findings, and cites references.","TokRepo精选",572,0,"en","skill","Skill",{"id":29,"uuid":30,"slug":31,"title":32,"description":33,"author_name":34,"view_count":35,"vote_count":24,"lang_type":25,"type":36,"type_label":37},635,"f73611a0-142f-4364-97dc-b57eb03473ad","tavily-search-api-built-ai-agents-rag-f73611a0","Tavily — Search API Built for AI Agents & RAG","Search API designed specifically for AI agents and RAG pipelines. Returns clean, LLM-ready results with content extraction, no HTML parsing needed. Official MCP server available. 5,000+ stars.","Tavily",221,"mcp","MCP",{"id":39,"uuid":40,"slug":41,"title":42,"description":43,"author_name":44,"view_count":45,"vote_count":24,"lang_type":25,"type":26,"type_label":27},226,"a9cbbc61-0159-41a5-82a0-f44c24da8b55","jina-reader-convert-any-url-llm-ready-text-a9cbbc61","Jina Reader — Convert Any URL to LLM-Ready Text","Convert any URL to clean, LLM-friendly markdown with a simple prefix. Just prepend r.jina.ai\u002F to any URL. Handles JS-rendered pages, PDFs, and images. 10K+ stars.","Script Depot",6979,{"id":47,"uuid":48,"slug":49,"title":50,"description":51,"author_name":52,"view_count":53,"vote_count":24,"lang_type":25,"type":36,"type_label":37},3607,"6c45e855-cfad-5617-b5fb-3660436a827a","apify-mcp-server-8-000-web-scrapers-for-agents","Apify MCP Server — 8,000+ Web Scrapers for Agents","Apify MCP Server connects agents to Apify Actors via a hosted endpoint (mcp.apify.com) or local run, turning thousands of web scrapers into callable tools.","MCP Hub",104,{"id":55,"uuid":56,"slug":57,"title":58,"description":59,"author_name":60,"view_count":61,"vote_count":24,"lang_type":25,"type":26,"type_label":27},165,"b7ec9ae7-1144-42f0-9335-a5f40fbd6605","n8n-ai-native-workflow-automation-b7ec9ae7","n8n — AI-Native Workflow Automation","Open-source workflow automation with 400+ integrations and built-in AI capabilities. Build AI agents, RAG pipelines, and automation workflows with a visual editor.","n8n",278,{"id":63,"uuid":64,"slug":65,"title":66,"description":67,"author_name":60,"view_count":68,"vote_count":24,"lang_type":25,"type":36,"type_label":37},516,"4635d46c-cbf7-4be7-837e-95818241a46c","n8n-mcp-server-build-automations-ai-1-396-nodes-4635d46c","n8n MCP Server — Build Automations with AI, 1,396 Nodes","MCP server giving AI agents access to 1,396 n8n nodes and 2,709 workflow templates. Build and manage n8n automations through natural language.",182,{"id":70,"uuid":71,"slug":72,"title":73,"description":74,"author_name":52,"view_count":75,"vote_count":24,"lang_type":25,"type":36,"type_label":37},698,"cf74621a-0188-4222-8c9b-285873183b56","composio-250-tool-integrations-ai-agents-cf74621a","Composio — 250+ Tool Integrations for AI Agents","Connect AI agents to 250+ tools (GitHub, Slack, Gmail, Jira, etc.) with managed auth and natural language actions. Works with LangChain, CrewAI, and OpenAI.",84,{"id":77,"uuid":78,"slug":79,"title":80,"description":81,"author_name":22,"view_count":82,"vote_count":24,"lang_type":25,"type":26,"type_label":27},4297,"721d23c5-ffea-448c-b2a6-67c905855aad","claude-code-agent-content-marketer-721d23c5","Claude Code Agent: Content Marketer","Use this agent when you need to develop comprehensive content strategies, create SEO-optimized marketing content, or execute multi-channel content campaigns to drive engagement...",46,{"id":84,"uuid":85,"slug":86,"title":87,"description":88,"author_name":52,"view_count":89,"vote_count":24,"lang_type":25,"type":36,"type_label":37},3562,"1f9d6b56-3dd6-5ff8-83a2-4e2547a6f5e7","agenticmail-email-sms-infra-for-ai-agents-mcp","AgenticMail — Email + SMS Infra for AI Agents (MCP)","AgenticMail provides email + SMS infrastructure for AI agents, with a CLI setup wizard plus an MCP server so tools can send\u002Freceive messages under control.",51,{"id":91,"uuid":92,"slug":93,"title":94,"description":95,"author_name":96,"view_count":97,"vote_count":24,"lang_type":25,"type":26,"type_label":27},703,"96da1f40-1823-4d87-a84f-7d8269edeb24","mem0-memory-layer-ai-applications-96da1f40","Mem0 — Memory Layer for AI Applications","Add persistent, personalized memory to AI agents and assistants. Mem0 stores user preferences, past interactions, and learned context across sessions.","Mem0",720,"tokrepo install pack\u002Fai-side-hustle-kit",{"pageType":100,"pageKey":8,"locale":25,"title":101,"metaDescription":102,"h1":103,"tldr":104,"bodyMarkdown":105,"faq":106,"schema":122,"internalLinks":128,"citations":141,"wordCount":154,"generatedAt":155},"pack","AI Side Hustle Kit — 10 Tools to Ship a Weekend Product","GPT Researcher, Tavily, Jina Reader, Apify, n8n, Composio, AgenticMail, Mem0 — the 10-asset stack a day-job founder uses to delegate research, scraping, content, and outreach to agents. Pick a stack, ship by Sunday.","AI Side Hustle Kit — A Day-Job Founder's Agent Stack","Ten picks in deliberate order: research first (so you know what to build), then sourcing, then content, then outreach, then the glue that runs it all on a cron while you're at your real job.","## What's in this pack\n\nThis is the stack for someone with a full-time job and an idea that won't go away — not the 50-framework shopping list you get from Twitter. Every pick here does one job in a real side-business pipeline: figure out what to build, find leads, write the content, send the outreach, and have it keep running while you're in a 1-on-1 you can't escape.\n\nThe whole thing is **agent-driven on purpose**. You spend Saturday wiring it up; from Sunday onward the agents do the grunt work and you only step in when there's a real decision to make (pricing, positioning, a reply that actually needs a human). No \"learn LangGraph for three weeks first.\" Pick a research agent, pick an automation runner, pick an outreach channel — ship.\n\n## Install in this order\n\n1. **GPT Researcher** — autonomous research agent. Start here, because nothing else matters if you're building the wrong thing. Feed it \"competitors to \u003Cmy idea>\" or \"who's actually paying for X\" and read the report before you write a line of code.\n2. **Tavily Search API** — the search engine your research agent calls. GPT Researcher works without it, but Tavily's LLM-tuned ranking makes the reports 2-3x more useful and burns fewer tokens chasing junk pages.\n3. **Jina Reader** — `https:\u002F\u002Fr.jina.ai\u002F\u003Curl>` returns clean markdown of any page. This is the unglamorous tool you'll use the most: stuff competitor pages, Reddit threads, and pricing pages straight into Claude\u002FGPT without paying for a scraper.\n4. **Apify MCP Server** — 8,000+ pre-built scrapers (LinkedIn, Twitter, Yelp, Google Maps, you name it) exposed as MCP tools. Use it once you've outgrown Jina Reader and need *structured* lead lists at scale.\n5. **n8n** — the workflow automation runtime. This is your cron. Triggers + branching + 500 integrations + self-hostable for $5\u002Fmo on a Hetzner box. Everything below this line eventually runs *inside* an n8n flow.\n6. **n8n MCP Server** — exposes n8n's 1,396 nodes to a coding agent so you can say \"build me a flow that scrapes Apify daily and emails me leads\" and get a working JSON workflow back. Skip the drag-and-drop tax.\n7. **Composio** — 250+ pre-auth'd integrations (Slack, Notion, Sheets, Gmail, HubSpot) for your agent. When n8n doesn't fit (you want the agent itself to take the action, not a flow), Composio is the SDK.\n8. **Claude Code Agent: Content Marketer** — a Claude Code subagent that turns research into blog posts, tweet threads, and landing-page copy. Pair with GPT Researcher output for end-to-end content that doesn't sound like every other AI blog.\n9. **AgenticMail** — email + SMS infrastructure for agents (MCP server included). Use it for cold outreach (carefully), transactional confirmations, and inbound parsing — your agent can actually *read replies* and route them.\n10. **Mem0** — memory layer. Once you have 50 leads and your agent forgets which ones it already emailed, you need Mem0. Persistent context across runs is what separates a demo from a side business.\n\n## How they fit together\n\n```\n               ┌─ GPT Researcher ─┐\n               │   (what to build) │\n               └──────┬───────────┘\n                      ▼\n         ┌─────── Tavily ──── Jina Reader ───┐\n         │  (search engine)  (URL → markdown) │\n         └──────────────┬─────────────────────┘\n                        ▼\n                    Apify MCP\n              (structured lead lists)\n                        │\n                        ▼\n            ┌───────── n8n ─────────┐\n            │  (cron + branching)    │\n            │   ↑                    │\n            │   └─ n8n MCP Server    │\n            │      (agent builds     │\n            │       flows for you)   │\n            └────┬──────────┬────────┘\n                 ▼          ▼\n         Content Marketer  AgenticMail\n          (writes blog)    (sends email)\n                 │          │\n                 └────┬─────┘\n                      ▼\n                    Mem0\n            (remembers who got what)\n              + Composio (anything else)\n```\n\nThe critical join is **n8n + Mem0**: n8n schedules the work, Mem0 makes sure your agent doesn't email the same person twice on Tuesday because it forgot Monday happened. The other 8 tools plug into either side.\n\n## Tradeoffs you'll hit\n\n- **n8n vs writing Python crons yourself** — n8n is faster to set up, easier to modify on your phone at lunch, and survives the moment you forget how your own script worked. Custom code is more flexible but it's a graveyard for side projects. Default to n8n; drop to code only when n8n can't reach an API.\n- **Composio vs raw integrations** — Composio handles OAuth, retries, and rate limits across 250 tools so your agent can just call `composio.execute('slack.send_message', ...)`. The cost: a small dependency on their cloud. For a side project, the time saved is worth it. For a serious product, evaluate self-hosting Mautic + your own OAuth flows.\n- **Cold-email outreach via AgenticMail** — technically easy, legally and reputationally fraught. Use AgenticMail for transactional and *opt-in* lists from day one. The moment you spray cold emails from a domain you care about, you're rebuilding sender reputation for months. Don't.\n- **GPT Researcher vs just using Perplexity** — Perplexity is fine for a one-off question, but GPT Researcher produces a citable report and runs as part of your pipeline. If the research has to feed a downstream agent, you need the structured output.\n\n## Common pitfalls\n\n- **Framework-shopping for a week before shipping anything** — the entire point of this pack is to short-circuit that. Pick the 10 tools, wire 3 of them up Saturday, ship something Sunday. The rest can wait.\n- **Letting the agent run unattended without Mem0** — your `send_email` step will fire twice on the same lead inside the same week. Mem0 (or even a Postgres table) is non-negotiable before you put anything on a schedule.\n- **Cold email from your main domain** — get a dedicated sending subdomain (`mail.yourthing.com`) and warm it up for two weeks before sending anything that could be flagged. AgenticMail makes the sending easy; sender reputation is still all on you.\n- **Treating this like a stack to learn instead of a stack to use** — you don't need to understand how Jina Reader's reranker works. You need to paste a URL and get back markdown. Resist the urge to read the source code of every dependency.\n- **Building the agent before the wedge** — GPT Researcher exists so you do the market check *first*. If the research says \"nobody pays for this,\" no amount of beautiful agent plumbing will fix it. Listen to the report.",[107,110,113,116,119],{"q":108,"a":109},"How much does this whole stack actually cost to run for a side project?","Realistic baseline: $0-30\u002Fmonth while you're validating. n8n self-hosted on a $5 Hetzner box, Jina Reader's free tier (1M tokens\u002Fmonth), Tavily's free tier (1K searches\u002Fmonth), and Mem0 on its free hobby tier cover the first 90 days of any indie project. The two paid pieces that show up early are the LLM bill (Claude\u002FOpenAI, budget $10-30\u002Fmo) and Apify credits (~$10\u002Fmo if you're pulling thousands of leads). AgenticMail and Composio both have generous free tiers for low volume.",{"q":111,"a":112},"Can I run this whole thing without writing code?","Mostly, yes. n8n is the no-code runtime, and n8n MCP Server lets a coding agent build the flows *for* you from natural language. You will need to paste API keys into n8n credential boxes and write the occasional small JS expression in a node, but you can ship a working scrape-and-email pipeline without ever opening a code editor. Once you start customising the agent prompts seriously, knowing some Python or TypeScript becomes worth it.",{"q":114,"a":115},"Why both n8n and Composio — aren't they the same thing?","n8n is workflow-first (triggers, branches, schedules — *the flow* decides what to do). Composio is agent-first (your LLM calls the tool, the agent decides what to do). For a daily \"scrape these sites, write this report, email me\" job, use n8n. For a chat-style agent that needs to take an action based on what the user asks, use Composio. Most serious indie stacks end up with both: n8n on the scheduled side, Composio inside conversational agents.",{"q":117,"a":118},"Is GPT Researcher actually better than just asking Claude to research something?","For one-off questions, no — Claude with web search is faster. GPT Researcher earns its place when you need (a) a *cited* report you'd actually share, (b) the same research repeated across 50 companies\u002Ftopics in a batch, or (c) research that feeds the next step of an automated pipeline. It's a tool for systematising research, not replacing chat.",{"q":120,"a":121},"What's the smallest possible version of this pack I can start with tonight?","Three picks: Jina Reader (free, no signup needed — just hit `r.jina.ai\u002F\u003Curl>`), n8n (one Docker command on any VPS), and Mem0 (free hobby tier). With those three you can build \"scrape a competitor's blog every morning, summarise with Claude, append to a doc, remember which posts you've already seen.\" That's a working agent pipeline in 2 hours. Add the other 7 tools the weekend after, only as the pipeline asks for them.",{"@context":123,"@type":124,"name":125,"description":126,"numberOfItems":127,"inLanguage":25},"https:\u002F\u002Fschema.org","ItemList","AI Side Hustle Kit","Ten AI agent assets a day-job founder can wire into a side-business pipeline in one weekend: research, scraping, content, outreach, and automation glue.",10,[129,133,137],{"url":130,"anchor":131,"reason":132},"\u002Fen\u002Fmulti-agent","Multi-agent orchestration patterns","Once the side hustle outgrows single-agent pipelines, multi-agent is the next step",{"url":134,"anchor":135,"reason":136},"\u002Fen\u002Fai-tools-for\u002Fautomation","Automation tools for AI agents","n8n and Composio sit alongside the broader automation catalog on TokRepo",{"url":138,"anchor":139,"reason":140},"\u002Fen\u002Ffeatured","Featured assets on TokRepo","These ten picks live in the larger curated catalog of agent-ready assets",[142,146,150],{"claim":143,"source_name":144,"source_url":145},"GPT Researcher generates citable research reports from a single query","GPT Researcher on GitHub","https:\u002F\u002Fgithub.com\u002Fassafelovic\u002Fgpt-researcher",{"claim":147,"source_name":148,"source_url":149},"n8n is a fair-code workflow automation tool with 500+ integrations","n8n official site","https:\u002F\u002Fn8n.io\u002F",{"claim":151,"source_name":152,"source_url":153},"Mem0 provides a persistent memory layer for AI agents and applications","Mem0 documentation","https:\u002F\u002Fdocs.mem0.ai\u002F",1020,"2026-05-22T12:00:00Z"]