All-in-one AI-powered B2B lead generation, automated website audit engine, cold outreach pipeline, and agency co-pilot.
Kashi Agency is a full-featured client acquisition platform built for agencies, freelancers, and B2B sales teams. It streamlines finding prospective clients, running automated UI/UX and SEO website audits, capturing real website screenshots, generating hyper-personalized cold emails, and closing deals using an AI Co-Pilot (Agent Kashi).
The application is engineered to work 100% free using standard free-tier API keys, and includes Offline Mode persistence so you can manage your CRM database, outreach logs, and deal pipelines even without an active internet connection.
- Client Finder: Search prospective leads by industry, location, and web presence.
- Pipeline Board: Drag-and-drop Kanban board (Not Contacted, Contacted, Negotiating, Closed Won, Closed Lost).
- Custom Client Management: Store detailed contact records, estimated revenue, website audit status, and notes.
- Real-Time Website Screenshots: Dual-tier screenshot fetching engine (
thum.io+ WordPressmshotsfallback) that loads clean, high-resolution website captures for any client URL. - AI Audit Scoring: Analyzes client sites for mobile responsiveness, page speed, CTA clarity, and SEO metrics.
- AI Task Automation: Delegate client research, email drafting, and pitch deck creation to your built-in agency assistant.
- Smart Prompts & Quick Actions: Generate audits, log team progress, or schedule follow-up emails in one click.
- AI Email Generator: Personalized cold outreach templates tailored to audit findings.
- Follow-up Automation: Schedule multi-touch email sequences and track open/reply rates.
- WhatsApp & Meeting Tracker: Log WhatsApp conversations and schedule client discovery calls.
- Video Walkthrough Simulation: Interactive video player demonstrating website audit teardowns for clients with real-time cursor navigation overlays.
- Media Vault: Organizes client audit files, screenshot previews, and shareable proposal links in structured folders.
- 100% Free API Compatibility: Operates seamlessly with free Google Gemini API keys.
- Offline Persistence: LocalStorage caching saves all lead updates, notes, email drafts, and outreach logs locally when working offline.
- Frontend: React 18, TypeScript, Vite
- Styling: Tailwind CSS, Lucide React Icons
- AI Engine:
@google/genai(Gemini 3.6 Flash) - Data Persistence: LocalStorage with automatic browser sync
- Deployment: Compatible with Vercel, Netlify, Cloud Run, and GitHub Pages
-
Clone the Repository
git clone https://github.com/YOUR_USERNAME/kashi-agency-crm.git cd kashi-agency-crm -
Install Dependencies
npm install
-
Set Up Environment Variables (Optional for AI Features) Create a
.envfile in the project root:GEMINI_API_KEY=your_free_gemini_api_key_here
You can get a free Gemini API key from Google AI Studio.
-
Start Development Server
npm run dev
Open
http://localhost:3000in your browser. -
Build for Production
npm run build
This project includes a pre-configured GitHub Actions Workflow (.github/workflows/deploy.yml) that automatically builds and publishes the web application to GitHub Pages whenever code is pushed.
- Go to your repository on GitHub:
https://github.com/engrwaqas-star/leadflow-ai-os - Click Settings -> Pages (in the left sidebar under Code and automation).
- Under Build and deployment -> Source, select GitHub Actions.
- Push any commit or trigger the workflow manually in the Actions tab. Your live app will be published at:
https://engrwaqas-star.github.io/leadflow-ai-os/
If you want to deploy from your terminal:
npm run deployThis builds the site and pushes the output directly to the gh-pages branch.
- Free Hosting: You can deploy this app for $0/month on:
- Vercel (
vercel deploy) - Netlify (
netlify deploy) - GitHub Pages (Static SPA build)
- Vercel (
- Free API Key: Obtain a standard Gemini API Key from Google AI Studio at no cost.
- Offline Working: When internet is disconnected, the top banner updates to Offline Mode Ready, preserving all changes locally in your browser!
Distributed under the MIT License. See LICENSE for details.