Project Info
Inspiration
“When we abandon facts, we abandon freedom.” — Timothy Snyder. Misinformation is lethal: 232 000 Americans died in 15 months after vaccine-hoax posts deterred them from a free COVID shot. 330 000 South Africans lost their lives—and 35 000 babies were needlessly infected—because one speech denied the HIV-AIDS link. The viral Cluely “cheat-on-everything” AI now pumps out polished essays, deepfakes, even fake lab data, eroding trust in all AI. Every new pandemic and every election is a bigger target. That’s the blaze—Truely is our fire-extinguisher. What It Does Truely is a plug-and-play browser extension that unleashes a fleet of AI agents the instant a page loads: How We Built It Frontend: WXT (Web Extension Toolkit) for modern Chrome extension development with TypeScript Svelte 5 for reactive, component-based UI with excellent performance TailwindCSS for utility-first styling and responsive design Backend: Chrome Extension Manifest V3 for secure background service workers and content scripts LETTA AI API for intelligent fact-checking and misinformation detection Anthropic Claude API as a secondary AI agent for robust analysis Custom DOM parsing engine to extract and clean webpage content for AI analysis WebExtension Messaging API for secure communication between extension components UUID-based action tracking for managing multiple fact-checking annotations Architecture Accomplishments We’re Proud Of Built a real-time AI fact-checker that flags misinformation directly in the browser Designed interactive annotations with clear, cited corrections that don’t interrupt reading Engineered a smart DOM pipeline that filters only essential user-visible content Shipped a performant and secure Manifest V3 Chrome Extension with advanced background capabilities Learnings & Challenges Parsing complex, inconsistent DOM structures across the web Compressing large webpage text into limited AI context windows without losing key facts Building a seamless system to highlight and annotate dynamic content without breaking page layout Coordinating responses across multiple AI agents with different output formats What’s Next Cross-page memory: Agents remember what you’ve seen and flag contradictions or recurring misinformation across sites. Custom AI agents, made by users: Anyone can deploy browser-native agents using just a prompt Monetization through agent marketplaces: Users can create, share, and even monetize their own agents via blockchain-powered rewards Truth has a new home address: your browser. We’re not just fighting misinformation—we’re making honesty the default experience on the internet. Join us.
Truely AI
An agentic browser extension to tell you the whole truth. See our devpost here.
Setup
We used WXT, Svelte and PNPM to build this project.
To get started, run pnpm install to install the dependencies and pnpm dev to start the development server.
Project Structure
This project uses WXT's standard project structure, with all source code located inside the src directory.
src/: Contains all the source code for the extension.src/assets/: This directory is for static assets that are processed by the build tool (Vite), like CSS, fonts, or images that are imported into your UI components.src/entrypoints/: This is where the main files for your extension live. Each subdirectory represents a different part of your extension.background/: The background service worker for the extension.content/: Content scripts that run in the context of web pages.sidepanel/: The sidepanel for the extension with configuration and a chat interface.
src/lib/components/: Contains reusable Svelte components that are used across different parts of the extension's UI.src/public/: Any files in this directory are copied directly to the root of the extension's build output without being processed.icon/: Contains the extension icons in various sizes.
wxt.config.ts: The main configuration file for the WXT framework, where you can customize the build process and extension manifest.package.json: The standard Node.js manifest file containing project metadata, dependencies, and scripts.
Recommended IDE Setup
VS Code with the following extensions:
Analysis
View
Metric
- 14
- 12
- 8
Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.
Technology
- AnthropicIn code
- CSSIn code
- SvelteIn code
- Tailwind CSSIn code
- TypeScriptIn code
5 of 5 appear in the indexed code.
AI coding agents
- CursorConfig
Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.
Codebase size
Source size
216 KB
Source files
22
Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.
Repository
nirpechuk/truely
36 files · 308 KB · @ ee92806
Structure
Interface
5 files · 14%Screens, components and styles rendered to the user.
Application logic
15 files · 42%Domain rules, services and shared utilities.
Supporting
Layers are inferred from where files sit in the tree, not from reading the code. A project that names its directories unconventionally will read oddly here — open the file browser to check anything the diagram implies.
Languages
- YAML73%
- TypeScript19%
- Svelte7%
- Markdown1%
- CSS0%
Share of indexed source by file size. Binary and vendored files are excluded.
Dependencies
package.json
npm · 26- @anthropic-ai/sdk
- @letta-ai/letta-client
- @webext-core/messaging
- dotenv
- mode-watcher
- runed
- uuid
- +19 more
Declared in the repository’s manifests at the indexed commit. A declared package is not proof it is used, and runtime dependencies are listed first.
This project’s features have not been analysed yet.
Export this project's context (description, README, evidence, key source files) to chat with an AI agent elsewhere.