Felipe de Leon

marketing engineer @ (our company)
i build the systems agencies claim they have

[Oliver Ye @_isomerism]

the moat is literally just caring about your work and being proud of your output and holy shit it does not seem to apply to 99% of people

[Felipe de Leon @felipedeleon]

@JamesonCamp doing this from a challenger position with low / forgotten expectations, its waaaay easier than doing it from a leading one. In a few days / weeks will be gone and X is a tiny bubble.

[Felipe de Leon @felipedeleon]

@elshayib_ feels like Cursor was able to break through B2B at bigger scale than other offerings out there and might be trying to cook something for use cases beyond code

[Felipe de Leon @felipedeleon]

Testing out and using Grok 4.7 paired with Hermes-Agent shows me that taste/ease-of-reading makes a huge difference for frontier labs.

[Felipe de Leon @felipedeleon]

@dhh @0xSero @0xSero well done! seems your last week tweet travelled you to denmark

[Felipe de Leon @felipedeleon]

@FardeemM Hermes-Agent, NousPortal. @NousResearch

[Felipe de Leon @felipedeleon]

@elshayib_ I had a old SOUL.md incentivizing autonomy that was causing these loopholes going in circles too much... I still like the speed tho

[Felipe de Leon @felipedeleon]

@martiano is there anyway to make it stretch across monitors?

[Felipe de Leon @felipedeleon]

@karan4d @yacineMTB Hermes Agent

[Felipe de Leon @felipedeleon]

sticking feathers up your butt does not make you a peacock.

[Felipe de Leon @felipedeleon]

@Teknium one thing isn’t very clear in long sessions (TUI) is when the model automatically changes to fallback OR at what point it did. I have fallback enabled, will disable to see what happens

[Felipe de Leon @felipedeleon]

have you already said fuck the zionists today?

[Le Grand Zack @Zack_Nani]

Et ben on ne va pas se gêner pour la partager du coup

[Felipe de Leon @felipedeleon]

@dara_venture damn. that's feel.

[Felipe de Leon @felipedeleon]

you also don't risk having your STT data sold/shared

[Felipe de Leon @felipedeleon]

you can just build your own. ask Hermes-Agent.

[Daniel Sempere Pico @dansemperepico]

WisprFlow is getting worse and worse each day for me.

Fortunately, it’s only a matter of time before WisprFlow is obsolete by native OS/app voice to text features.

But that time cannot come soon enough.

[Felipe de Leon @felipedeleon]

my number 1 tip is to use Hermes-Agent from @NousResearch and not face this problem at all.

[lauren @poteto]

my number 1 tip for controlling your grok @bot usage is to avoid scheduled routines that run too frequently. for example, a 15 min routine runs almost 100 times a day, and every run consumes tokens. hourly or a few times a day is usually good enough.

the length of the chat with your bot can also make routines much more expensive. for recurring ones, try giving that to a fresh bot, while you continue your chat with your main bots (like a chief of staff).

give this post to your bot and ask them to help!

[Felipe de Leon @felipedeleon]

@ivanfioravanti similar here, but with the ultra plan from @NousResearch

[Felipe de Leon @felipedeleon]

@witcheer in addition you can declare the exact token amount with 'compression.threshold_tokens', which defers back to ratio-based if the value is higher than models context window. I prefer when switching between 1M and 200k context models.

[Felipe de Leon @felipedeleon]

Cancelled all my AI services. Doubling-down on @NousResearch Ultra subscription.

a) never limited by usage.
b) models + tools + vm
c) rollover when credits not used.
d) minimum I can do to support its mission.

you should as well.