Skip to content

ROADMAP.md

Latest commit

 

History

History
194 lines (132 loc) · 5.16 KB

ROADMAP.md

File metadata and controls

194 lines (132 loc) · 5.16 KB

LearnHub – Roadmap

This roadmap is aspirational, not a strict promise.
It exists to show where LearnHub is headed and how new ideas fit into the bigger picture.

Status labels:

  • ✅ Shipped
  • ⚙️ In progress
  • 🔜 Planned / Next
  • 💡 Idea / Later

1. Recently shipped

✅ AI-powered buttons on every summary

  • For each summary.*.md, automatically inject AI buttons such as:
    • Teach (Beginner / Intermediate / Advanced)
    • Analogy
    • Cheat Sheet
    • Mindmap
    • Flashcards
    • Practical Projects
    • Code Examples
    • Common Mistakes
    • Quiz
    • Interview Me
    • Assessment Rubric

Each button:

  • builds a clear, human-readable prompt,
  • wires the summary URL and fallback links into the AI chat bot,
  • and sends the user directly to their chosen provider.

✅ Multiple fallback URLs for summaries

  • When generating a summary, we now upload it to:
    • alisol.ir
    • at least one paste service (e.g. paste.rs)
    • optionally another fallback like hastebin
    • and sometimes the original source

This increases the chance that the chosen AI model can actually fetch the context.

✅ Per-request teaching flows

  • Different flows now exist for:
    • teaching at different levels (beginner / intermediate / advanced),
    • generating flashcards with CSV export,
    • creating analogies, mindmaps, quizzes, interview simulations, and rubrics.

✅ Ethical disclaimer support

  • Every teaching flow can inject a disclaimer like:

    “DISCLAIMER: It’s not a substitute for the original material.
    Please study the main source from <instructor_name> if you can: <source_url>.”

2. Now / In progress

⚙️ Standardising summary header and structure

  • Ensure each summary.*.md includes:
    • AI-powered button block,
    • clear metadata (resource type, instructor name, main source URL, tags),
    • consistent sections for the summary itself.

Goal: make summaries:

  • easy to read,
  • predictable for learners,
  • and reliable as context for AI models.

⚙️ Repository documentation

  • README.md improvements
  • VISION.md (this document)
  • ROADMAP.md (you are here)

3. Next

🔜 Provider selection UI improvements

  • Show only providers that support:

    • query-parameter prompts, and
    • external URL fetch (where possible).

  • Display simple capability hints, for example:

    • “Good for coding”
    • “Good for fresh web knowledge”
    • “Can generate downloadable files (CSV, etc.)”
    • “Better with long prompts”

  • Encourage users to:

    • use a subscribed / paid tier if possible,
    • and choose a strong enough model for tutoring.

🔜 Task-based model recommendations

For each workflow, LearnHub should guide the user to the best types of models:

  • Flashcards:

    • suggest models that can produce downloadable CSV files (e.g. better for Anki).
    • hide or de-prioritise models that cannot generate files.

  • Heavy coding / complex backend tasks:

    • highlight models that are strong at code generation and reasoning.

  • Extra-fresh knowledge / outdated sources:

    • highlight models that combine strong web search with reasoning.

Implementation idea:

  • A small, human-readable capability matrix in code and docs.
  • Future: auto-generated from a config file.

🔜 Contribution flow: “Request a resource”

Because summaries are not crowd-sourced, the main contribution path will be:

  • A simple issue template:
    • “Request a topic or resource”
    • fields: topic, your level, preferred language/stack, suggested resource links (optional)

The maintainer will:

  • check if the resource is suitable,
  • confirm whether LearnHub already covers it,
  • and, if it fits, add it to the internal “to-summarise” queue.

4. Later / Ideas

💡 Open-source backend

  • Open-source the backend that:
    • generates prompts,
    • constructs URLs,
    • uploads summaries to multiple hosts,
    • and handles aliases/fallbacks.

Benefits:

  • full transparency around prompts,
  • easier external contributions (bug fixes, new workflows),
  • and more trust from users and instructors.

💡 Better capability awareness

  • Show per-provider limitations, for example:

    • which models can generate files,
    • which ones struggle with external URLs,
    • which ones are best for long conversations vs small tasks.

  • Autodetect when a model replies:

    • “I can’t open URLs right now…”
      and suggest switching to another provider.

💡 Smarter routing

In the long term, LearnHub could:

  • analyse user intent (flashcards, deep dive, interview prep, etc.),
  • and suggest a default recommended model and prompt variant,
  • while still letting the user stay in full control.

💡 More learning workflows

Examples:

  • spaced repetition helpers,
  • topic roadmaps (multi-resource learning paths),
  • “compare two resources” flows.

5. Non-goals (for now)

To stay focused, LearnHub does not plan to:

  • host or stream full video courses,
  • sell access to other people’s educational content,
  • become a generic marketplace or SaaS product.

The core focus is:

High-quality human summaries + smart AI workflows
to help serious learners study existing resources better and faster.