Gordon FontenotRelated to my recent string of vinyl posts, I made an Obsidian plugin that makes it easy to generate detailed notes for a music release last week. It’s not available in the official registry yet (approval is backed up by months right now) but you can always unzip it into your vault manually. I’m using it along with Obsidian Bases to keep a nice catalog of records I own (including where/when I bought them) and which albums I’m currently looking for. I think it’s neat!
Justin StachBuilt Markset: tiny native macOS app that renders Markdown with Finder's QuickLook (press Space). No app to open, no file associations to change. Free.
If you live in .md files: Obsidian, READMEs, notes, I'd love to know what breaks.
Gauff 🇪🇺Information overload alert 😪
After 3 tears of using #obsidian #obsidianmd , I wanted to try something lighter and #opensource (important to me). I did some #joplin , some #logseq ...
I have too many notes where I didn't put a hashtag to categorise the information, didn't migrate everything from one to another... So my stuff is all over the place, not organised, and it drives me nuts 🙁
At the end of the day, I don't even know which one I'd use. Probably #logseq . maybe I could script something to dump useful content of other apps onto a big note in Logseq, so that a full text find works? #1bigtextfile @ellane ?
Or go to the shop and buy a notebook!
fossilesqueGitHub - domleca/llm-wiki
leeleedee@passthejoe I think #obsidian can be as complicated (or as simple) as one wants it to be.
I have used it for years now and love it, both for my work - where it is my primary project management/support ticket tracker/CRM tool for my freelance business - and for organising my hobbies, especially the #DnD campaign I run.
In those years I have tweaked and customised my set up of it so much that an outsider looking at it might think it's super complicated but it works perfectly for me and my various needs.
However when I first started my set up was super simple and for a long time I was only using just the basic notes formatted with the inbuildt #markdown language, which took me about an hour to learn. It was literally just notes with links to other notes if needed. I didn't even have folders.
And that's the beauty of #obsidianmd you can customise it as much/as little as you want, and make it into whatever you need it to be.
Yes, I am an obsidian fangirl, and I make no apologies 😄🤣
Magnus Hedemark@jcrabapple I'm "learning in the open", BTW, so not only are most of my notes public... but there's a lot documented in there about how I approach note keeping in #ObsidianMD, how I'm evolving through different memory substrates for the agent, etc.
https://publish.obsidian.md/magnus919/8+-+Topics/_Topic+Tree
I've got a sort-of agent running in #ClaudeCode that I really love but I want to depend less on the #Anthropic ecosystem. I don't mind running foundation models, but I want to be able to change up what models I am using.
#OpenClaw is certainly an option. But a bit "Wild West" right now.
#HermesAgent has been interesting. I think my first stab at it could have gone better. But then I leaped in without really reading any of the docs. I'm thinking I might spin up another one just for the purpose of going through the docs and taking the time to learn it, concept by concept, before more thoughtfully designing the real agent I want.
The Claude Code agent that I'm using has a few things going for it that I'd like to port over (or improve upon):
- Molecular note keeping partner using #ObsidianMD. I've been at this awhile so my use is pretty sophisticated and I've had a hard time getting smaller local models to play along well.
- Agent memory based on #LadybugDB, tracking promises, semantic spacetime vertices & edges... This has been pretty cool and I think I want to recreate it (better) on #ArcadeDB.
- Also using #LightRAG as my hybrid RAG platform. I'd want to migrate this to #ArcadeDB, as well.
- I think I'm going to create more sophisticated MCP tools that are more thoughtfully designed for context management and containment of tool proliferation.
- I'd love to get to a place where I could use a Hermes Agent harness for most of my work, local models for a lot of smaller work (I've only got a 16GB VRAM GPU) and foundation models (or #OpenRouter) for the heavy lifting.
Random GeekHm. Different Obsidian CLI commands support different output formats ("format=json|tsv|csv", "format=json|yaml|tsv", no format option at all for this one, etc).
The `obsidian plugin id=<name>` command offers no structured format and provides less information than just dumping `.obsidian/plugins/<name>/manifest.json`
Gonna be honest: even for a proof of concept this evokes caution.
Having entirely too much fun with Bases and themes and datetime puzzle solving.
wephI released a new version of my Bookshelf plugin for Obsidian yesterday.
If you're into reading, managing your books (or book notes), tracking your reading progress, and checking stats, why not give it a try?
Plugin: https://obsidian.md/plugins?id=bookshelf
Docs: https://weph.github.io/obsidian-bookshelf/
Code: https://github.com/weph/obsidian-bookshelf