Skip to content

AIFrontiersLab/IdeaTap

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TrendCanvas ✨

Turn LinkedIn ideas into polished, export-ready visuals — instantly.

Build Platform Swift License PRs Welcome

TrendCanvas (formerly IdeaTap) helps creators, founders, recruiters, and professionals transform a single copied LinkedIn link into high-quality visual assets—posters, carousels, and clean export-ready graphics—tuned to the original content’s intent and enhanced with timely AI news + trends.


Why TrendCanvas exists (problem)

LinkedIn is full of strong ideas—great posts, profiles, frameworks, and insights. But turning those ideas into polished visuals usually means:

  • 🕒 Spending 30–120 minutes in design tools
  • 🧠 Translating a post’s “intent” into a visual concept
  • 📰 Manually adding timely context (news, trends, recent releases)
  • 🧩 Iterating formats (single image vs carousel) and export specs

Most people give up—or post low-effort assets that don’t match the quality of the idea.


What TrendCanvas does (solution)

TrendCanvas makes “visualizing” a LinkedIn idea as easy as:

  1. 🔗 Copy a LinkedIn post or profile link
  2. 🧠 Let the AI agent infer intent + audience + tone
  3. 🎨 Generate a tuned visual set (poster, carousel, cover images)
  4. 📦 Export clean, social-ready assets in one click

Core product vision

TrendCanvas is the fastest path from “idea” → “shareable visual.”
It’s built for a modern creator workflow: minimal friction, high taste, consistent branding, and relevance that updates with the world.


Key features (highlights)

  • 📋 Automatic LinkedIn URL detection from the system clipboard
  • 🧾 Context capture (post/profile content + inferred themes)
  • 🧠 AI intent analysis (audience, tone, goal, CTA, format)
  • 🖼️ Tuned visual generation (posters, carousels, clean layouts)
  • 📰 Trend enrichment using the latest AI news, insights, and zeitgeist
  • One-click asset generation built for speed
  • Export-ready outputs (clean typography, spacing, and contrast)
  • 🧼 Premium minimal aesthetic suitable for demos + investor decks

Screenshots (placeholders)

Replace these with real images when available.

App

  • Home / Clipboard detection
    • docs/screenshots/home.png
  • Link preview + intent
    • docs/screenshots/intent.png
  • Generated carousel preview
    • docs/screenshots/carousel.png
  • Export options
    • docs/screenshots/export.png

Sample outputs

  • docs/samples/poster_01.png
  • docs/samples/carousel_01.png

Example user journey (from copy → export)

Step What you do What TrendCanvas does
1 Copy a LinkedIn post/profile link Detects clipboard change + validates URL 🔎
2 Click Generate Fetches/derives context, infers intent, picks best format 🧠
3 Pick a style preset Applies a consistent visual system (type, grid, colors) 🎛️
4 Review assets Shows poster + carousel variants with tuned captions 👀
5 Export Exports social-ready images (and future: templates) 📦

Real-world use cases

  • Creators: turn a saved post into a carousel with your voice + better structure
  • Founders: convert an announcement into a clean launch visual set
  • Recruiters: transform a profile or hiring post into brand-consistent hiring graphics
  • Professionals: turn a thread into a “framework poster” for easy sharing
  • Communities/teams: generate weekly trend recaps with up-to-date AI news

AI workflow & agent architecture

TrendCanvas uses an agentic pipeline that separates capture → understanding → enrichment → design → export.

High-level flow (diagram)

User copies LinkedIn URL
        │
        ▼
Clipboard Listener (on-device)
        │
        ▼
URL Validator + Link Router
        │
        ├── Post URL ───────────────┐
        └── Profile URL ────────────┤
                                    ▼
                         Context Capture Layer
                     (metadata + extracted text)
                                    │
                                    ▼
                           Intent Analyzer Agent
               (audience, tone, goal, key points, CTA)
                                    │
                                    ▼
                          Trend Enrichment Agent
        (latest AI news, releases, trends, insights; optional)
                                    │
                                    ▼
                           Visual Composer Agent
          (layout system, typography, hierarchy, variants)
                                    │
                                    ▼
                        Renderer + Export Packager
                (poster, carousel, sizes, filenames)
                                    │
                                    ▼
                               Share / Save

Agents (roles)

Agent Responsibility Output
Intent Analyzer Understand “what this wants to achieve” structured brief (tone, audience, bullets, CTA)
Trend Enricher Add timely relevance (AI news + insights) enrichment notes + citations/links (optional)
Visual Composer Convert brief into design-ready content blocks layout spec + copy blocks + variant plan
Renderer Produce exportable assets PNG/JPG (and future: editable templates)

Tech stack (placeholder)

Update this section once finalized.

  • Client: Swift / SwiftUI (iOS, macOS)
  • Clipboard capture: native clipboard APIs
  • AI orchestration: TBD (agent framework, prompts, tools)
  • Visual rendering: TBD (CoreGraphics / SwiftUI rendering / server-side)
  • News/trend sources: TBD (curated feeds / APIs)

Installation

Prerequisites

  • macOS with Xcode installed
  • Swift toolchain (via Xcode)

Build & run

  1. Clone the repo
  2. Open IdeaTap.xcodeproj in Xcode
  3. Select a target device/simulator
  4. Run the app (⌘R)

Environment variables (placeholder)

Add your keys using your preferred secrets workflow (e.g., Xcode build settings, .xcconfig, Keychain, or secure on-device storage).

Variable Description Required
TRENDCANVAS_AI_API_KEY AI provider key ✅/⬜
TRENDCANVAS_NEWS_API_KEY Trend/news provider key ✅/⬜
TRENDCANVAS_ENV dev / prod

Usage

  1. Open TrendCanvas
  2. Copy a LinkedIn post URL or profile URL
  3. Confirm the detected link inside the app
  4. Choose a preset (tone/style)
  5. Generate visuals
  6. Export and share

Project structure

Current structure (may evolve as agents/rendering expand).

IdeaTap/
  IdeaTap.xcodeproj/          # Xcode project
  IdeaTap/                    # App source
    ContentView.swift
    IdeaTapApp.swift
    LinkSheetService.swift
    Assets.xcassets/
  IdeaTapTests/
  IdeaTapUITests/
docs/                         # (recommended) screenshots, samples, diagrams

Roadmap

Near-term

  • Add robust LinkedIn context extraction (post text, author info, themes)
  • Style presets + brand kits (type scale, colors, logo placement)
  • Carousel generation with consistent grid + pacing
  • Better export controls (sizes, safe areas, filenames)

Mid-term

  • Trend enrichment with citations + “why it matters”
  • Templates library (frameworks, announcements, hiring, recaps)
  • Multi-platform export presets (LinkedIn, X, Instagram)

Long-term

  • Team workspaces + shared brand guidelines
  • Editable template exports (e.g., Figma/Canva-compatible) (TBD)
  • Scheduled “weekly trend recap” auto-generation (opt-in)

Privacy & security

TrendCanvas is designed to be privacy-forward:

  • 🔐 Clipboard access is used to detect copied LinkedIn URLs and should be minimized and transparent.
  • 🧭 Scope-limited capture: only process content needed to generate the requested assets.
  • 🗑️ Data retention: default to ephemeral processing where possible; avoid storing raw content unless the user saves a project.
  • 🧩 Secrets handling: API keys should never be committed; store securely via platform mechanisms.

This project is early-stage. Please treat privacy/security as a first-class requirement when contributing.


Contributing

Contributions are welcome. If you’d like to help:

  1. Fork the repo
  2. Create a feature branch (feature/your-change)
  3. Make changes with clear commits
  4. Open a PR with:
    • what you changed
    • how to test it
    • screenshots (if UI)

If you’re adding AI behaviors:

  • Keep prompts/tooling deterministic where possible
  • Prefer structured outputs (JSON-like schemas) for agent handoffs
  • Include safety checks for URLs and user-provided content

License

TBD — add your preferred license (e.g., MIT, Apache-2.0) and update the badge at the top.

About

IdeaTap lets users click copy on LinkedIn, auto-capture profile URLs, then an AI agent creates tuned visuals with the latest AI news insights.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages