$Basecat holder fked up
-30% in a bull market
One of the worst pain holders going through
I'm averaging here making big money or complete fk up, taking that risk
last cycle:
- i turned $16k into $20m+ on $BONK
- i turned $6k into $1m+ on $WIF
- i turned low 6 figures into $8m+ on $FARTCOIN
- i turned low 5/6 figures into $1m+ each on $POPCAT, $PNUT, etc.
- i turned low 5/6 figures into $5m+ each across several narratives, including the PolitiFi meta (MAGA, DOGEGOV) and Roaring Kitty meta (AMC, GME, KITTY, etc.)
every single one of these trades was taken and chronicled publicly while the majority of the timeline ridiculed me and said they would never play out (just search my X and you'll see it all there!)
yet i think many are still grossly underprepared for what is about to happen with $USELESS
i genuinely think this could end up being by far my biggest memecoin trade
$USELESS is by far the strongest memecoin narrative i've encountered, and the confluence of metrics i pay attention to โ the same metrics that helped me identify all of the trades above โ is pointing toward one conclusion:
$USELESS is shaping up to be THE memecoin play of this bull run
most people still aren't positioned for what i think $USELESS is about to do
they will be.
Hermes Agent hit 2,500 contributors sometime this weekend.
Amazing what the community can do coming together, and we've barely scratched the surface on reviewing and integrating everyone who's contributed's work!
Thank you all for making Hermes Agent great!
@baseapp From past 1 week i been struggling to get back my funds back that i sent to robinhood chain,
Then In wallet it is saying robinhood chain isn't supported lol i thought any eth chain is available there is no custom chai add function,
@jessepollak@cobie plz help me to get my funds
GPT-5.6 is dropping in the next hour.
If youโre going to try it for coding, donโt make the obvious mistake:
Do not pay the smartest model to reread your repo from zero.
That is the most expensive way to use it.
The trick is simple:
use GPT-5.6 for reasoning,
not repo archaeology.
Most people will open a fresh session and paste:
- project summary
- same files again
- same architecture notes again
- same โwhat changedโ context again
- same instructions from the last session again
That burns premium context before the model has done any real work.
For the last 25 days, weโve been building around one idea:
what if the agent could recall the repo instead of rereading it?
That became Memory Recall.
It turns your repo into local, governed memory for Codex, Claude Code, Cursor, and other coding agents.
The workflow is:
1. map the repo
2. store reviewed repo facts locally
3. track changed-file impact
4. generate a compact handoff
5. send the model only what it needs
6. keep proof of what was actually sent
On our own repo, we measured:
- 72% fewer delivered tokens on session delta delivery
- 84.6% fewer delivered tokens in Rust graph-query eval
- 84.35% smaller practical context-pack report
These are delivery-token measurements, not provider billing claims.
But the practical lesson is obvious:
If GPT-5.6 is expensive, stop using it as a search engine for your own codebase.
Use it for:
- hard debugging
- architecture decisions
- code review
- execution planning
- reasoning across the right files
Let memory handle the repeated context.
Install:
npm install -g memory-recall
Then:
recall setup
recall handoff
No hosted account.
No model API key.
No silent memory capture.
Make GPT-5.6 recall your repo instead of rereading it.
@jun_song Bro you won my heart โค๏ธ,do not listen to all these guys,keep on building those who are telling you about do not involve in crypto tokens,these guys are already in crypto I saw that guy replied, you can see bitcoin in his bio these guys want to fud and try to get entry on token