> For the complete documentation index, see [llms.txt](https://ai-os-and-trend-finder.gitbook.io/ai-os-and-trend-finder-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://ai-os-and-trend-finder.gitbook.io/ai-os-and-trend-finder-docs/docs/hackathon/hackathon-project-brief.md). # Hackathon Project Brief: Trend Finder This document preserves the competition brief and the current local demo proof path. The early sections mention "dashboard or HTML report" because that was the competition prompt; the implemented Trend Finder MVP is the local dashboard with Brief, Watchlist, and Engine Replay surfaces, not a generated HTML report. ## Project **Name:** Trend Finder\ **Context:** May Competition\ **Issued by:** Jack Roberts / AI Automators Community\ **Submission deadline:** 26 May 2026\ **Required submission:** 60-second Loom walkthrough plus a GitHub repository, posted under `May Comp` ## Goal Build a dashboard or HTML report that finds trending AI topics before they become mainstream. Trend Finder should help AI creators, automators, and builders identify high-signal topics early enough to turn them into useful content, experiments, or product ideas. The project should separate noise from signal by collecting public trend data, ranking it, and presenting the strongest opportunities in a clear, usable format. ## Mission Create a working dashboard or HTML report that: * Finds trending AI topics from public sources * Surfaces hidden gems before they become obvious * Shows enough supporting evidence for each trend * Helps a creator decide what to make content about next * Turns raw internet activity into a ranked, readable report ## Why This Matters Unique content performs better than generic content. For AI creators, the hard part is not only finding what is already popular, but spotting what is starting to move before everyone else is talking about it. Trend Finder is intended to make that discovery process faster and more reliable by combining signals from places where AI topics emerge early. ## Audience The primary users are: * AI automators * Content creators * Solo builders * Newsletter writers * YouTube creators * Developer advocates * Product researchers These users need fast answers to questions like: * What AI topics are gaining momentum right now? * Which tools, repos, workflows, or debates are still under the radar? * Why is this topic worth paying attention to? * What content angle could I create from this trend? ## Possible Data Sources The project can use any sources that help identify AI trend signals, including: * Reddit discussions * YouTube videos and comments * GitHub repositories and star velocity * Hacker News posts * Product Hunt launches * AI newsletters * Research papers * Model release notes * Developer blogs * X/Twitter or other social platforms, where available The strongest version of the project should combine multiple sources so trends are not based on a single weak signal. ## Suggested Dashboard or Report Sections A useful Trend Finder report could include: * **Top Trending AI Topics:** the strongest topics ranked by current momentum * **Hidden Gems:** lower-volume topics showing early acceleration * **Why It Is Trending:** short explanation of the signal behind each topic * **Evidence:** source links, post counts, star growth, comment activity, or video velocity * **Trend Score:** a simple scoring model for ranking opportunities * **Creator Angle:** suggested content ideas, hooks, or audience questions * **Source Breakdown:** where each trend is appearing * **Watchlist:** topics that are not ready yet but may become important soon ## Success Criteria The project should be judged by how well it: * Finds genuinely useful AI topics * Identifies early signals, not only obvious mainstream trends * Explains why each trend matters * Provides evidence from real sources * Presents findings in a clean dashboard or HTML report * Helps creators make faster content decisions * Demonstrates a repeatable process that can be run again * Feels creative, practical, and useful within the hackathon timeframe ## Deliverables The final submission should include: * A GitHub repository with the project code and setup instructions * A dashboard or generated HTML report * A 60-second Loom video explaining the build * A post under `May Comp` The Loom should quickly explain: * What the project does * Which sources it uses * How trends are detected or ranked * What makes the output useful * One or two example trends discovered by the system ## Dashboard Demo Proof Points The implemented MVP delivery surface is the local dashboard, not a generated HTML report. Use [Trend Finder Demo Workflow](/ai-os-and-trend-finder-docs/docs/hackathon/trend-finder-demo.md) before recording to choose either the credentialed reviewed-source path or the credential-free fixture/demo path. Use [Hackathon Submission](/ai-os-and-trend-finder-docs/docs/hackathon/hackathon-submission.md) as the final Loom script and validated proof checklist. A 60-second Loom can use this path: 1. Open `http://127.0.0.1:5189/extensions/trend-finder/trends`. 2. Show Runtime readiness and provenance labels to explain the current AI provider state, credential presence, data origin, and analysis state without exposing secrets. 3. Show the Creator Lens and Run Trend Finder or Save and run control only as a local dev-server aggregate refresh, not a hosted job. 4. Show Source Health to identify active, degraded, offline, and blocked sources. 5. Open a ranked trend card and point to its score breakdown: momentum, novelty, evidence strength, source diversity, niche fit, and creator potential. 6. Open the source context and evidence links. Show that each topic is grounded in source URLs, browser-safe public metric chips when present, and missing references surfaced instead of hidden. 7. Highlight movement or previous-score labels when the payload includes them. 8. Highlight creator angle, suggested hooks, and audience questions as the creator-facing output. 9. Navigate to Hidden Gems for early lower-volume topics. 10. Navigate to Watchlist for generated monitoring rows with movement or reason labels. 11. Navigate to Brief to show the compact creator handoff: top creator angles, evidence to verify, and source health. Use one of these accurate proof-point sets. Credentialed reviewed-source run: * Say configured reviewed public sources, not unrestricted live source collection. * Show `Live generated data` only if the dashboard displays that label. * Show `AI-analyzed run` only if the dashboard displays that label. * If a source is degraded or blocked, call out the label instead of hiding it. Credential-free fixture/demo run: * Say fixture/demo data and deterministic fallback analysis. * Mention that no live AI clustering is claimed in this path. * Show disabled runtime and blocked/degraded source labels if present. Keep the demo language scoped to what exists today: local dashboard runtime, configured reviewed public sources, evidence-backed scoring, creator angles, dashboard Brief handoff, and visible fallback behavior. Hosted deployment, dynamic source marketplace, unrestricted source collection, and generated HTML report delivery are future work. Final overclaim guardrails: * Do not claim all reviewed sources run live by default. * Do not claim fixture/demo topics are live source output. * Do not call deterministic fallback output live AI clustering. * Do not present degraded or blocked source states as active sources. * Do not claim official OpenAI API or Agents SDK analysis paths are implemented. * Do not call the local run control a hosted job, scheduler, or marketplace. * Do not claim Creator Lens edits rescore the already-rendered payload; they affect output after save and rerun. ## Competition Prizes * **1st place:** $500 * **Most Creative:** $300 * **Random entries:** 2 mystery prizes ## Optional Teaming Participants can work solo or partner with someone else. A strong team could split the work between data collection, ranking logic, dashboard design, and presentation. ## Positioning Statement Trend Finder helps AI creators discover early AI topic signals and turn them into timely, differentiated content opportunities. --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://ai-os-and-trend-finder.gitbook.io/ai-os-and-trend-finder-docs/docs/hackathon/hackathon-project-brief.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.