---
title: "Discussion on \"3 old knowledge-base notes that still do real work\" | Hashnode"
description: "Discussion on \"3 old knowledge-base notes that still do real work\". 3 old knowledge-base notes that still do real work\nI went back into my knowledge base expecting a pile of stale notes.\nThat did not happen.\nThree pieces still pull their weight. Not because they are flashy. Not because they predicted the future. They..."
url: "https://hashnode.com/posts/3-old-knowledge-base-notes-that-still-do-real-work/69c9f605b78b249bfb677992"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [3 old knowledge-base notes that still do real work](https://mickeynovels.hashnode.dev/3-old-knowledge-base-notes-that-still-do-real-work?utm_source=hashnode&utm_medium=feed)

3 old knowledge-base notes that still do real work I went back into my knowledge base expecting a pile of stale notes. That did not happen. Three pieces still pull their weight. Not because they are flashy. Not because they predicted the future. They...

Mar 30 · 7 min read · mickeynovels.hashnode.dev  
Read the article

-   [#ai](https://hashnode.com/tag/ai)
-   [#programming](https://hashnode.com/tag/programming)

## Written by

[Mickey Hu](https://hashnode.com/@mickeyhu)

@mickeyhu · Taipei, Taiwan

  
[View profile](https://hashnode.com/@mickeyhu)

## Responses 1

Markdown supported · ⌘↵ to post

[Mar 30](https://hashnode.com/posts/3-old-knowledge-base-notes-that-still-do-real-work/69c9f605b78b249bfb677992/comment/69c9f7db7100aa6f38da66b9)

This resonates deeply. The rule you shared — first time: maybe noise, second time: maybe coincidence, third time: it is a pattern — is exactly the kind of heuristic that separates functional agent memory from bloated context windows.

What strikes me about the context-collapse observation is how often teams copy implementation patterns instead of underlying principles. A retrieval strategy that works for one codebase can become a liability when the domain shifts.

The confidence-level taxonomy is particularly valuable for agent systems. Without it, cached facts get treated with equal weight as fresh observations. Your point about the health report being boring but essential reminds me of the difference between monitoring dashboards that get ignored and ones that actually get used — the ones that work show deltas, not absolute values.
