How It Works

The Challenge

Hacker News discussions aren't simple comment lists — they're deeply nested, branching conversations where context matters. A reply five levels deep might reference something from the original post, and the most insightful comment might be buried in a side thread. Traditional summarization tools miss these nuances entirely.

Thread-Aware Representation

HN Companion preserves the hierarchical structure of conversations using a path-based notation plus engagement signals. Each comment keeps its parent/child relationship and includes score, reply count, and downvotes, so the AI understands not just what was said, but where it sits and how strongly the community engaged with it.

[1] (score: 1000) <replies: 3> {downvotes: 0} user1: Main point as the first reply to the post
[1.1] (score: 800) <replies: 1> {downvotes: 0} user2: Supporting argument or counter point in response to [1]
[1.1.1] (score: 150) <replies: 0> {downvotes: 6} user3: Additional detail as response to [1.1], but should be excluded due to more than 4 downvotes
[2] (score: 400) <replies: 1> {downvotes: 0} user4: Comment with a theme different from [1]
[2.1] (score: 250) <replies: 0> {downvotes: 1} user2: Counter point to [2], by previous user2, but should have lower priority due to low score and 1 downvote
[3] (score: 200) <replies: 0> {downvotes: 0} user5: Another top-level comment with a different perspective

This representation lets the AI track branches, identify sub-discussions, and prioritize quality signals while de-emphasizing or excluding low-signal comments.

What We Capture

Our summarization is designed specifically for technical discussions. The AI is prompted to:

• Identify main topics — Extract the key themes and arguments being discussed
• Capture diverse viewpoints — Surface different perspectives and notable opinions, not just the most upvoted takes
• Track conversation shifts — Note where discussions pivot to new topics or important tangents emerge
• Include relevant quotes — Pull brief, impactful quotes with references back to the original comments
• Maintain objectivity — Present information neutrally without editorializing

Smart Prioritization

Not all branches are equal. Before summarization, comments are sorted by relevance and engagement, with the most active conversation branches weighted higher. This means the AI focuses on what the community found most interesting, while still capturing valuable insights from quieter threads.

Adaptive Summaries

Summary length scales with thread complexity. A quick 20-comment discussion gets a brief overview, while a sprawling 500-comment debate receives a more comprehensive breakdown. The goal is always to give you just enough context to decide what's worth reading in full.

The Result

Instead of generic bullet points, you get summaries that actually reflect how HN discussions work:

• Context preserved — replies make sense without reading parents
• Side discussions surfaced — valuable tangents don't get lost
• Debate captured — disagreements and different viewpoints included
• Actionable — jump directly to interesting comments from the summary