---
title: "Discussion on \"Why I rebuilt my AI second brain — three problems that surfaced after five weeks\" | Hashnode"
description: "Discussion on \"Why I rebuilt my AI second brain — three problems that surfaced after five weeks\". Five weeks after v1.0, three problems showed up in my Obsidian + Claude Code second brain. v2.0 is the rebuild — skills + agents, per-area Projects, Standards."
url: "https://hashnode.com/posts/rebuild-ai-second-brain-architecture-structure-standards/6a1e743042e1d5f57cb05b9d"
---

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

Discussion

# [Why I rebuilt my AI second brain — three problems that surfaced after five weeks](https://blog.deepakgoyal.ai/rebuild-ai-second-brain-architecture-structure-standards?utm_source=hashnode&utm_medium=feed)

Five weeks after v1.0, three problems showed up in my Obsidian + Claude Code second brain. v2.0 is the rebuild — skills + agents, per-area Projects, Standards.

Jun 2 · 13 min read · blog.deepakgoyal.ai  
Read the article

-   [#obsidian](https://hashnode.com/tag/obsidian)
-   [#claude-code](https://hashnode.com/tag/claude-code)
-   [#secondbrain](https://hashnode.com/tag/secondbrain)
-   [#pkm](https://hashnode.com/tag/pkm)

## Written by

[Deepak Goyal](https://hashnode.com/@deepakgoyal-ai)

@deepakgoyal-ai · Faridabad, India

  
[View profile](https://hashnode.com/@deepakgoyal-ai)

Technical Architect · AI-powered systems at scale · Architecture decisions, not opinions

Technical Architect · 20 years in software · Writing about AI architecture, system design, and real decisions at scale

## Responses 3

Markdown supported · ⌘↵ to post

[Adam Lewis](https://hashnode.com/@nark3d)

[Jun 7](https://hashnode.com/posts/rebuild-ai-second-brain-architecture-structure-standards/6a1e743042e1d5f57cb05b9d/comment/6a25a8671a98581b1dd0814c)

We've seen the same pattern on smaller solo projects too. Authoring skill N+3 three weeks after skill N-1 with no PR review is exactly the failure mode.

We'd copy the 110-line cap straight across. A number the agent can't talk past does more enforcement work than a written rule.

Out of interest - how did you settle on 110 specifically?

[Deepak Goyal](https://hashnode.com/@deepakgoyal-ai)

[Jun 14](https://hashnode.com/posts/rebuild-ai-second-brain-architecture-structure-standards/6a1e743042e1d5f57cb05b9d/comment/6a25a8671a98581b1dd0814c#reply-6a2e39f313a2695c8a3616b3)

Hi Adam,

Honestly, it came from observation more than theory.

After building SMRITI, skills were changing almost daily — each iteration adding a few lines here, a new step there. Token usage kept climbing because every invocation loads the full skill file. At the same time, skills that grew too fast became harder to reason about.

We looked at where all the existing skills actually landed and 100–110 was the natural sweet spot — tight enough to keep costs predictable, generous enough that nothing needed an emergency rewrite. So we just drew the line there.

Thanks Deepak

[Adam Lewis](https://hashnode.com/@nark3d)

[Jun 14](https://hashnode.com/posts/rebuild-ai-second-brain-architecture-structure-standards/6a1e743042e1d5f57cb05b9d/comment/6a25a8671a98581b1dd0814c#reply-6a2ed0ca4977fe82ab813440)

That makes sense, an observed cap is more defensible than one picked up front because it comes from what the skills actually needed. And the number matters less than the fact that hitting it makes you stop and check whether the skill has quietly taken on a second job. Cheers for the detail.