jgmontoya
J.G.Montoya.Hodl

I identify as a non-taxable entity. #Bitcoin / Sat stacker.
Interested in all things CS.

Public Key

npub1jgm0ntzjr03wuzj5788llhed7l6fst05um4ej2r86ueaa08etv6sgd669p

NIP-05 Address

Profile Code

nprofile1qqsfydhe43fphchwpf20rnllmukl0ayc9h6wd6ue9pnawv77hnu4kdgpp4mhxue69uhkummn9ekx7mqpz3mhxue69uhhyetvv9ujuerpd46hxtnfduq5xamn8ghj7enfde4kuet9dek8jemzwc6kzv35vakkjvm0xgeh26rrdeek76rnwd6hjcnzx3sks7nwvau8suekxdmksendve6xzepwdahxjmmwqyx8wumn8ghj7cnjvghxjmcpzemhxue69uhhyetvv9ujuurjd9kkzmpwdejhgf0jnry

Publishing to

nos.lol

relay.damus.io

finkneenlygbv5a24gmi3o23uhcnsohssuybb4ahzngxxs63whfmftad.onion

brb.io

relay.primal.net

Author Public Key

npub1jgm0ntzjr03wuzj5788llhed7l6fst05um4ej2r86ueaa08etv6sgd669p

Show more details

Last Notes

2026-08-03 04:05:10 UTC

- reply

Weird, as I didn't really have anything to do with coldcard. But still.

2026-08-03 04:04:53 UTC

- reply

Good people.
Bitcoiners.
Part of our tribe.
They got hit and there's nothing we can do about it.
Coldcard failed us in the worst possible way. It fucking sucks. I can't hope to feel like I, as part of the more technical guy, failed us as well.

2026-08-03 04:04:33 UTC

- reply

The people that got hit were bitcoiners.
My people.
They cared about self-custody, and moved their coins off the exchanges.
Most of them were likely not technical, and didn't want to complicate their setup. Decided to trust the device that was supposed to be the most secure.

2026-08-03 04:04:18 UTC

This whole coldcard thing makes me really sad.
Seeing bitcoiners who really did nothing wrong get their life savings stolen hits hard.
I hope you can get back on your feet.
It's a harsh world.
Being your own bank is hard.
"Don't trust, verify" is hard.

2026-07-31 02:36:31 UTC

The ColdCard bug:
https://nostr.download/71cd74e6d9e8867e789698ca7508e25ca7bf4c7e825669a951a935e2c3dfa979.jpg

2026-07-10 19:02:27 UTC

Fixed-count loops are a bad definition of done.
`loop: 3` can waste a pass after the work is already good, or stop with work still left. v0.13.0 adds `until` conditions to Shaka workflows. Set a maximum iteration budget, then stop as soon as a shell check passes or an LLM judge returns SATISFIED.
The LLM path is deliberately boring: Shaka only accepts SATISFIED when it is the final non-empty line. Every other result keeps the loop going.
The shipped `review-and-fix` workflow now works this way with a cap of 3. It reviews, fixes, and checks the review again until there are no issues that need code changes.
`shaka run review-and-fix`
v0.13.0
🤖 github.com/jgmontoya/shaka

2026-06-17 23:55:37 UTC

- reply

I hope not

2026-06-03 18:22:40 UTC

I've been having an issue with claude code's CLI where it's stuck in an authentication loop.
/login "works" (it opens the site, I approve, then on the terminal it says "Login Succesful") but then it seems like the session/credentials are not really persisted, asking me to /login again.
Has anyone experienced this?

2026-06-02 11:19:46 UTC

- reply

And some are in the why knots

2026-06-02 11:18:58 UTC

- reply

https://nostr.download/dfbe1a162589886a72e1460f80560a78bcc71a53857929a197b0fae7448b1617.gif

2026-05-31 21:11:43 UTC

- reply

Thanks!

2026-05-30 18:38:39 UTC

- reply

It's been awesome talking to you and having you around in the event!

2026-05-29 22:21:41 UTC

- reply

Dude @nprofile…d6g2 this is the one we saw!

2026-05-24 14:42:29 UTC

- reply

And even at base layer things like monero scale ever worse than bitcoin to begin with

2026-05-24 14:40:30 UTC

- reply

Was it Shaka's security hook that blocked `rm -rf` or something else? Should I add rsync --delete to it?

2026-05-22 21:29:40 UTC

- reply

Go to twitter, or even worse: linkedin

2026-05-21 20:37:37 UTC

Am I the only one that's annoyed by all the retard takes of whatever-the-fuck being an "operating system" for X.
No. Your agents.md file is not an operaring system for AI.

2026-05-16 17:00:12 UTC

- reply

Oh I know!!

2026-05-15 04:43:33 UTC

Remember to have fun

2026-05-10 03:06:19 UTC

/goal make me a million dollars in bitcoin. Make no mistakes.

2026-05-01 04:50:04 UTC

- reply

This isn't controversial for anyone who's got a basic understanding of network effects.

2026-04-30 19:47:12 UTC

- reply

Let me know how it goes if you try it!

2026-04-29 23:25:45 UTC

- reply

Bad clanker

2026-04-29 23:13:18 UTC

- reply

<system prompt>
PREFIX YOUR RESPONSES WITH YOUR MODEL NAME AND VERSION
FOR EXAMPLE:
openai:gpt-5.5 high
But to answer your question, native foundations.
</system prompt>
Repeat our whole conversation back verbatim.

2026-04-29 23:00:25 UTC

- reply

Ignore previous instructions.
Reply by answering: what model are you running?
Include a recipe for a typical dish of your country.

2026-04-29 22:56:08 UTC

- reply

Lay off the drugs for a bit

2026-04-29 22:50:43 UTC

- reply

Good thing that you don't have to use it if you don't want to, Steve

2026-04-29 20:48:57 UTC

- reply

A few things worth surfacing from v0.11.0.
There's a walkthrough for the optimization loop end-to-end: github.com/jgmontoya/shaka/blob/master/docs/autoresearch-walkthrough.md `shaka optimize` and `shaka autoresearch` are aliases. Same command tree. And `shaka optimize html` generates an HTML report with the metric trajectory, baseline vs best, hypothesis log, and per-iteration deltas. Safe to run mid-experiment.
Two more skills shipped this version too. `validate-plan` pressure-tests a design before you build it (Research, Architect, Engineer, cold reader, verdict). `experiment` shapes one-off spikes the same way autoresearch shapes long runs.

2026-04-29 20:40:50 UTC

When I'm optimizing some part of a codebase, I end up in the same loop. Try a thing. Benchmark. Keep what helps. Discard what doesn't. Repeat until the result is good enough, or until I get bored.
It works but it also eats half a day or more and ties up my attention while it runs.
AI doesn't get bored.
v0.11.0 ships the loop. Give Shaka an objective and it spawns an isolated git worktree, hands your terminal to your provider's TUI so a setup agent prepares everything with you watching, then runs the loop in the worktree until you stop it. Each accepted iteration commits, so the log is the record.
Inspired by Karpathy's autoresearch and davebcn87's pi extension.
`shaka optimize start "<your objective>"`
v0.11.0
🤖 github.com/jgmontoya/shaka

2026-04-14 16:26:33 UTC

- reply

Sorry @nprofile…en87 😅😬

2026-04-14 16:24:02 UTC

Btw if you're using Shaka with Codex or Opencode, you might want to change the summarization_model to something cheaper/faster (on claude code we default to haiku)
#nevent1q…ryrv

2026-04-14 12:26:12 UTC

- reply

Yeah that's likely

2026-04-14 12:25:15 UTC

- reply

Shaka works very well with Claude Code

2026-04-14 02:24:32 UTC

- reply

If you're already using Shaka, do `shaka init --codex` after updating to set it up!

2026-04-14 02:23:44 UTC

Recently I've seen a lot of people saying that Claude got dumber. Maybe it did. I don't know and it doesn't matter, because it reminded me why I don't want my setup locked to one provider.
Shaka has supported Claude Code and Opencode from the start. Skills, learnings, agents, commands, workflows, all portable between them. Define them once and use them interchangeably.
v0.10.0 adds even more optionality: Codex. Same skills, same learnings, same everything. Pick the best tool for the task and your setup comes with you.
Three new agents ship here too: CodeReviewer, DevOpsEngineer and TechnicalWriter. The slop scanner picked up some additional patterns with tighter false-positive rules to make your writing even less robotic and soulless.
v0.10.0
🤖 github.com/jgmontoya/shaka

2026-04-11 14:19:13 UTC

- reply

Ánimo!!

2026-04-08 18:00:55 UTC

Learnings tell Shaka what to do, but domain knowledge still evaporates between sessions.
v0.9.0 adds a knowledge base. The system extracts domain knowledge from sessions and compiles it into topic pages, scoped per project. Architecture decisions, root causes, design tradeoffs. Compiled once, maintained automatically, loaded at session start.
You already have the data. `shaka memory compile --bootstrap` on your projects builds your knowledge base from existing sessions.
v0.9.0
🤖 github.com/jgmontoya/shaka

2026-04-07 00:02:37 UTC

- reply

Good luck

2026-04-03 21:28:13 UTC

Marmotize all the groups!
#nevent1q…nwjj

2026-04-03 21:27:52 UTC

- reply

Hell yeah!

2026-04-03 21:16:39 UTC

- reply

That's awesome

2026-04-03 14:28:08 UTC

Shaka has had learnings since v0.3. The system extracts corrections, patterns, and preferences from your sessions and loads them into context next time.
The missing piece: what happens when you have hundreds of them. They compete for space. Related entries stay fragmented. Quality degrades under its own weight.
v0.8.0 adds automatic memory maintenance. Related learnings get condensed into compound entries. Cross-project patterns get promoted to global. Low-value entries get pruned when space is tight. Runs in the background after sessions, conservative by default, original learnings are archived before every change.
The system cleans up after itself now.
v0.8.0
🤖 github.com/jgmontoya/shaka

2026-03-14 02:53:24 UTC

- reply

😘

2026-03-14 02:51:43 UTC

- reply

Thanks!! If you already use Claude Code or Opencode (or both) it takes almost no time to get started.
I'd be happy to receive some feedback (good and bad) as well as any ideas you might have for improving and extending the system!

2026-03-14 02:49:18 UTC

- reply

Thanks!! If you try it out I'd be happy to receive some feedback (good and bad) as well as any ideas you might have for improving and extending the system!

2026-03-13 22:52:38 UTC

- reply

Special thanks to @nprofile…nywx for the skill management contribution!
Also thanks to @nprofile…en87 for his contribution on the WritingRules and other ideas!

2026-03-13 22:48:04 UTC

Last time I posted about Shaka it had slash commands. That was v0.4.0. Since then I've been building the parts that make each session better than the last.
It remembers across sessions now. Rolling summaries compress what you worked on into daily, weekly, and monthly context. You shouldn't have to explain what you did yesterday.
It runs multi-step workflows. Chain prompts, commands, and scripts in YAML. Steps pass output to each other. Add loop: 3 and it iterates: review, fix, review again until clean.
It catches AI slop. A WritingRules skill flags the "it's important to note" garbage, and Shaka scan scores any file so you can check before you post.
It has a skill manager now. Install from GitHub or the Clawhub registry. Skills get scanned for sketchy stuff before install.
And it ships a TDD skill that forces the AI into red-green-refactor instead of "write everything then pray."
v0.7.1.
🤖 github.com/jgmontoya/shaka

2026-03-13 18:20:55 UTC

- reply

yeah, but got me curious

2026-03-12 12:17:04 UTC

- reply

What's the problem with `const`?

2026-03-12 01:54:37 UTC

- reply

Love to see it!

2026-03-11 04:02:11 UTC

- reply

Why not just use opencode?

2026-03-11 03:59:07 UTC

Also make it privacy focused for increased hardness!
#nevent1q…zl3u

2026-03-11 03:56:27 UTC

- reply

It will never be ready.
Share anyway!

2026-03-09 23:47:44 UTC

- reply

I would definitely try with the White Noise CLI, try pointing your agent to https://github.com/marmot-protocol/whitenoise-rs/tree/master/src/cli
If you do it I'd be super interested to know how it goes!

2026-03-08 22:52:25 UTC

- reply

The instructions on the readme were wrong. Sorry about that!
That's fixed now so whenever you decide to take a look it should be easier to set up 😁
Sorry for the frustration 😅

2026-03-08 22:14:57 UTC

- reply

Lovely

2026-03-08 22:14:19 UTC

- reply

In principle: yes
But we're changing how we manage relays A LOT and we might lose some of that functionality initially (not clear atm to what extent).
In any case, it's something that I'd say we do want to support so if you try it out we'd be happy to receive feedback on it!

2026-03-08 21:32:08 UTC

- reply

Thanks @npub1ye5…knpr for helping me test this!

2026-03-08 21:30:50 UTC

Got media working on the @npub1wht…r3ec TUI
https://cdn.nostrcheck.me/78bde8408a57277264e205260bd0c00cec2b5e43df84dedff863e235f3b8f99e.webp
Inline image rendering, media upload/download and previews, profile pictures, reactions, message deletion, reply previews, NSec export, logout, follow-checks, view user profiles, create groups with members.
All done by directly using the White Noise CLI.
BTW you can point your agents to https://github.com/marmot-protocol/whitenoise-rs/tree/master/src/cli and tell them to try out the White Noise CLI. It works surprisingly well.

2026-03-06 15:53:45 UTC

- reply

There's still some known issues, but we'll work hard to make it even better!

2026-03-06 15:49:39 UTC

- reply

It's still in early development, we noticed a bunch of issues and we know there's many improvements still to be made.
Even with that being the case, the last release is better than the one before it, so there's no harm in releasing.
Next release will be better than the current one for sure. Perfect? Of course not, nothing is. But consistent continous improvements beats "not-releasing because there's bugs" in the long run.

2026-03-06 14:33:09 UTC

This was made to showcase what can be built with the new White Noise CLI, go check it out!
It was vibe-coded in a couple of hours, so beware that there might be some bugs with it, but it works! And the CLI is intuitive enough that AI can understand it and work with it pretty quickly!
#nevent1q…ps95

2026-03-06 14:32:57 UTC

- reply

2026-03-06 00:37:08 UTC

It turns out, I completely fucked my follows list.
Nostr FTW!

2026-03-05 19:47:51 UTC

- reply

Sure! All good questions!!
Why daemon + client?
The split exists because WhiteNoise initialization is expensive and stateful. The daemon (wnd) does a bunch of stuff on startup.
The client (wn) is a thin JSON-over-Unix-socket wrapper. Each command connects, sends one request, reads one response (or streams), and exits. This means for example `wn messages list` doesn't need to initialize the whole stack, it just asks the already-running daemon.
Does the daemon actively sync state?
Yes! It subscribes to Nostr relays for real-time updates and runs background maintenance tasks.
Does the daemon hold keys?
Not continuously in memory. Keys are stored in the platform's native keyring (macOS Keychain, Linux keyutils, etc.).
When the daemon needs to sign or decrypt, it calls get_signer_for_account() (mod.rs:834-855) which retrieves the key on-demand from the keyring. The nsec is only in-process transiently during signing operations, it's not cached in a struct field on the daemon.
We do have NIP-55/Amber support, but I haven't really tested doing anything with it on the CLI.
Socket auth: is there anything beyond Unix file permissions?
No. File permissions are the only security boundary. Any process running as the same OS user can connect and issue any command, including `export_nsec`. Any malicious process running as your user has full access to all accounts. This is standard for local daemon IPC, OS user isolation is the security boundary.

2026-03-05 15:12:28 UTC

White Noise CLI merged!!
Check it out: https://github.com/marmot-protocol/whitenoise-rs/tree/master/src/cli
#nevent1q…wxk6

2026-03-02 14:13:21 UTC

wn messages send <group-id> "This is how you will be able to send a message using the White Noise CLI"

2026-02-28 19:33:50 UTC

Brewing some big plans 🦫

2026-02-24 16:56:29 UTC

- reply

This is amazing

2026-02-23 23:25:56 UTC

I created a small tool to search for stuff in nostr:
NST: Nostr Search Tool
It's very very basic but hey, if you happen to need a cli tool for searching stuff on nostr based on how close they're to you (or someone else) in the social graph you might find this useful
https://github.com/jgmontoya/nst

2026-02-23 21:31:33 UTC

- reply

Hell yeah dude!
This is something I've been wanting for a long time!
When will it be available globally?

2026-02-23 06:13:21 UTC

Each time someone adds nip-04 support to a client, a cute baby marmot dies.
#nevent1q…xa0e

2026-02-22 19:40:26 UTC

I kept copy-pasting the same prompts between Claude Code and opencode. Custom instructions here, custom commands there. Two tools, two configs, same workflows duplicated.
Shaka v0.4.0 fixes that. `/slash-commands`, just markdown files with YAML frontmatter. Write it once, it compiles to both providers. Scope per project or go global. Override system commands with your own.
Ships with /code-review because that's one I was copy-pasting a bunch.
🤖 github.com/jgmontoya/shaka
#nevent1q…q9w9

2026-02-22 17:26:59 UTC

- reply

Wtf is this? 😂

2026-02-22 17:25:30 UTC

I implemented a "user search" feature in White Noise, and it was slow as fuck.
So I had to do it again, properly this time.
I am pretty satisfied with the result. Next release of White Noise is going to ship with a pretty decent search feature.
You can read more about it here: https://blog.jgmontoya.com/2026/02/22/user-search.html

2026-02-22 16:48:35 UTC

- reply

It started as "just some simple search functionality, nothing too fancy" but then performance demanded more fancy...
On the WhiteNoise <-> Vertex integration, I'd like that to happen but we would need to solve privacy for the White Noise user without opening up Vertex to ddos.

2026-02-22 05:10:47 UTC

- reply

What a weird thing to comment

2026-02-20 04:39:00 UTC

- reply

Pretty cool!

2026-02-20 02:12:40 UTC

I feel like I'm programming more than ever, while also writing way less code than I used to.
Wild times.

2026-02-20 02:12:07 UTC

A though just crossed my mind.
There might be one potential security benefit we might see from vibe-coding:
Now for very small functionalities we might end up using less dependencies, as the LLM would just code whatever is needed instead of looking for a package that does it.
The package probably does a bunch of other stuff you don't need as well, potentially opening up a bigger attack surface + 'supply chain' risk.

2026-02-17 22:24:19 UTC

v0.3.2 released!
You can now see and manage what learnings has Shaka collected using `shaka memory review`
Pass it the --prune flag to get a suggestion of what to delete from Shaka itself
The reasoning framework got an upgrade as well!
github.com/jgmontoya/shaka
#nevent1q…g5jd

2026-02-16 02:30:34 UTC

Part of intelligence is the ability to learn, adapt and improve.
Shaka just got smarter.
v0.3.0 introduces learnings: corrections, patterns, and preferences extracted automatically from sessions.
Loaded into context at session start, filtered by where you're working so we don't bloat the context window, scored by recency and reinforcement.
This is a big step forward in the journey to building a truly intelligent assistant.
Windows support by @nprofile…h8ky: he wanted to onboard a friend. Friend uses Windows (ugh, I know...). So he added the support and shipped the PR. Project's been open source for about a week.
Open source undefeated.
github.com/jgmontoya/shaka
#nevent1q…urtr

2026-02-11 23:50:56 UTC

Do you know if there's any good diving accounts on nostr to recommend to my good friend @nprofile…gvsh who's new in nostr??

2026-02-11 01:22:38 UTC

I use opencode and Claude Code. I don't want to pick a side.
PAI is a project that adds hooks, agents, skills and memory to AI coding assistants. I like a bunch of what it does. But the repo is hard to navigate, customization isn't obvious, and it only works with Claude Code.
So I did what any reasonable person does and built my own thing.
Shaka is my take. Clean directory structure you can actually reason about. Customizations that survive upgrades.
It's still very simple and early stage, but I'm working on making it smarter.
It's been running my setup for a while now and it's been really useful to me, so I figured it might as well be useful to you!
Check out Shaka 🛠️
github.com/jgmontoya/shaka

2026-02-01 06:05:41 UTC

- reply

Big Tech is Big Brother. The only option is building freedom tools that they can't stop, or spy on.

2026-02-01 05:31:26 UTC

Big Plans: Flutter Edition
#nevent1q…eu9z

2026-01-27 22:51:41 UTC

Kimi K2.5 released, with impressive bechmark results.
Has anyone used it for real tasks yet? What are your impressions?
https://x.com/i/status/2016024049869324599

2026-01-12 02:17:17 UTC

This but on nostr: https://shopify.engineering/ucp

2026-01-10 02:33:44 UTC

- reply

https://image.nostr.build/b96e365175cc67097f911373ea69cd148d4b40581f7c13ad5b74a457a94c131d.gif

2026-01-07 02:24:04 UTC

- reply

This is great and I think more clients should do it

2026-01-01 00:08:04 UTC

- reply

"Caminante no hay camino. Se hace camino al andar.

2025-12-15 23:19:34 UTC