kepano's Avatar

kepano

@stephango.com

making @obsidian.md

7,171
Followers
747
Following
402
Posts
25.04.2023
Joined
Posts Following

Latest posts by kepano @stephango.com

Preview
GitHub - kepano/defuddle: Extract the main content from web pages. Extract the main content from web pages. Contribute to kepano/defuddle development by creating an account on GitHub.

yes, that's exactly what defuddle is
github.com/kepano/defud...

04.03.2026 22:31 πŸ‘ 5 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
Obsidian Web Clipper Highlight and capture web pages in your favorite browser. Save anything and everything with just one click.

I made Defuddle for Obsidian Web Clipper
obsidian.md/clipper

04.03.2026 17:36 πŸ‘ 9 πŸ” 1 πŸ’¬ 1 πŸ“Œ 0

yes :)

04.03.2026 16:06 πŸ‘ 4 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I love the way the artifacts look in this picture

04.03.2026 16:06 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Video thumbnail

Defuddle now has a website!

This means you can use Defuddle anywhere to get the main content of a page in Markdown format.

You can simply add "defuddle.md" before any URL, use it via curl, Skills, CLI, or add it to your app via NPM.

04.03.2026 16:02 πŸ‘ 213 πŸ” 26 πŸ’¬ 9 πŸ“Œ 2

correct

27.02.2026 23:35 πŸ‘ 8 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

eta is now
bsky.app/profile/obsi...

27.02.2026 21:55 πŸ‘ 10 πŸ” 1 πŸ’¬ 2 πŸ“Œ 0

Anything you can do in Obsidian you can do from the command line.

27.02.2026 16:25 πŸ‘ 131 πŸ” 3 πŸ’¬ 5 πŸ“Œ 0

taste is a study of nuance, it requires slowing down

18.02.2026 19:31 πŸ‘ 49 πŸ” 5 πŸ’¬ 0 πŸ“Œ 0

that's a beautiful thought, thanks for sharing it

13.02.2026 13:51 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

It's not because your agent can do something with 12 tool calls and 20k tokens that it should.

For example, Obsidian CLI has a command called "orphans" that finds files with no incoming links.

Now your agent can get a deterministic answer in 1 millisecond instead of 4 minutes.

11.02.2026 21:09 πŸ‘ 77 πŸ” 6 πŸ’¬ 1 πŸ“Œ 2

Obsidian uses local files, so all your standard terminal commands work for editing/moving/searching/etc

Obsidian CLI adds everything else:

- interacting with the UI
- internal functions e.g. base queries, orphans, etc
- devtools, console, screenshots, eval, etc

See the CLI docs!

10.02.2026 18:15 πŸ‘ 116 πŸ” 14 πŸ’¬ 3 πŸ“Œ 0

You can define a specific view with ![[Base#View]]

06.02.2026 04:54 πŸ‘ 3 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

it's working... slowly πŸ˜†

06.02.2026 04:53 πŸ‘ 3 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Post image

A hearty oatmeal and buckwheat chocolate chip cookie for winter time
stephango.com/hearty-choco...

01.02.2026 20:22 πŸ‘ 49 πŸ” 1 πŸ’¬ 3 πŸ“Œ 2

exo = external so exocortex may be more apt than exobrain

you are still the pilot

20.01.2026 22:16 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Video thumbnail

I have never liked the term "second brain" because it sounds externalized and passive

ideally Obsidian should feel like strapping into your exosuit, a direct extension of yourself

20.01.2026 18:51 πŸ‘ 121 πŸ” 7 πŸ’¬ 6 πŸ“Œ 1

I made the same assumption! Clearly I am living in a filter bubble of my own making. Thank you Dan!

19.01.2026 01:02 πŸ‘ 5 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

@danabra.mov would you mind putting a little link to file over app in your excellent piece about formats over apps?

18.01.2026 17:23 πŸ‘ 50 πŸ” 1 πŸ’¬ 4 πŸ“Œ 0

ATProto brings @stephango.com's "file over app" philosophy to the network.

18.01.2026 16:15 πŸ‘ 41 πŸ” 5 πŸ’¬ 0 πŸ“Œ 1

social file over social app

18.01.2026 17:08 πŸ‘ 19 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Preview
GitHub - mProjectsCode/obsidian-bases-charts-plugin Contribute to mProjectsCode/obsidian-bases-charts-plugin development by creating an account on GitHub.

github.com/mProjectsCod...

17.01.2026 20:02 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

If you're enjoying Obsidian 1.11 consider giving us a review on the App Store or Google Play.

As a small team you have no idea how much motivation it gives us to keep improving the app πŸ₯°

15.01.2026 18:25 πŸ‘ 41 πŸ” 6 πŸ’¬ 3 πŸ“Œ 0

1. Are you still able to scroll up in the base to see the content?

2. Load time change is unrelated to the update, it's because you're opening to a base which requires loading the entire vault... if you open to a note that won't happen. You can change the default note to open in Settings β†’ Editor

15.01.2026 14:45 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

Try the default theme

15.01.2026 05:18 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

you can already do that
bsky.app/profile/step...

14.01.2026 14:42 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
Private inference When you use an AI service, you’re handing over your thoughts in plaintext. The operator stores them, trains on them, and–inevitably–will monetize them. You get a response; they get everything.

I like this blog post, I think it lays a solid foundation that many more AI companies should adopt
confer.to/blog/2026/01...

13.01.2026 17:24 πŸ‘ 23 πŸ” 1 πŸ’¬ 0 πŸ“Œ 0
Confidential computing
This is the domain of confidential computing. Confidential computing uses hardware-enforced isolation to run code in a Trusted Execution Environment (TEE). The host machine provides CPU, memory, and power, but cannot access the TEE’s memory or execution state.

LLMs are fundamentally statelessβ€”input in, output outβ€”which makes them ideal for this environment. For Confer, we run inference inside a confidential VM. Your prompts are encrypted from your device directly into the TEE using Noise Pipes, processed there, and responses are encrypted back. The host never sees plaintext.

But this raises an obvious concern: even if we have encrypted pipes in and out of an encrypted environment, it really matters what is running inside that environment. The client needs assurance that the code running is actually doing what it claims.

Confidential computing This is the domain of confidential computing. Confidential computing uses hardware-enforced isolation to run code in a Trusted Execution Environment (TEE). The host machine provides CPU, memory, and power, but cannot access the TEE’s memory or execution state. LLMs are fundamentally statelessβ€”input in, output outβ€”which makes them ideal for this environment. For Confer, we run inference inside a confidential VM. Your prompts are encrypted from your device directly into the TEE using Noise Pipes, processed there, and responses are encrypted back. The host never sees plaintext. But this raises an obvious concern: even if we have encrypted pipes in and out of an encrypted environment, it really matters what is running inside that environment. The client needs assurance that the code running is actually doing what it claims.

As hyped as Claude Code + Obsidian is, I have no desire to upload my personal unencrypted vault data to Anthropic or anyone else.

I hope that the ideas of private inference and confidential computing that Moxie described take off. It's how all LLMs should work.

13.01.2026 17:24 πŸ‘ 132 πŸ” 9 πŸ’¬ 4 πŸ“Œ 2

I agree, but in the meantime try Command palette β†’ Change theme

12.01.2026 21:46 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

yes

12.01.2026 21:45 πŸ‘ 6 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0