@_nnaci

https://xitter.cf/t.co/bkPzBp0Dln

 bangeristan
Joined July 2024
  • Tweets122
  • Following66
  • Followers219
7Media
jmazzola.github.io/blog/muse… new blog post is up on muse wonder full's RE capabilities, what surprised me, what it sucks at, and how much meta truly cooked with this model

meta's muse wonder full solved flare-on 13, then wrote its own blog post about it

no face scan, no id, no refusals. meta's muse wonder full solved flare-on 13 (8/9 solo, one with hints) in under eight hours on a $50 plan, then wrote a blog post selling itself.

jmazzola.github.io
3
1
9
243
1
3
117
lemme drop some free alpha (omp)
3
4
2
56
2,548
gave astra a try again, very beautiful
42
6
6
526
68,288
cheers
dword64@dwordxyzOct 15
Naci's "readable" code passed every check.
 He rejected it anyway because it was still ugly. Cheers~
2
78
naci retweeted
i²cjak
@i2cjakOct 14
Chud labs is announcing our new decision model, Chud-1 - Distilled down to a single 'false' return - 100% on Nothing Ever Happens bench - The West has fallen
Ethan Lambeth
@LambethEthanOct 14
Coin labs is announcing our new decision model Flip-1. - Distilled down to 1 parameter - 50% on yes/no bench - Runs fully offline
30
122
7
3,114
43,050
your computer after you go into recovery mode, hit omp with the “pls fix”, and walk away:
10
4
159
5,146
naci retweeted
naci@_nnaciOct 13
Wrote up my attempt at deobfuscating Luraph-protected Lua: lifting Lua to LLVM IR , then using AI + partial evaluation to strip the VM, and checking it against the original. nac-l.github.io/2026/10/12/d…

Deobfuscating Luraph: Lifting Lua to LLVM, Then Letting AI Try

Lifting Lua to LLVM IR, why dynamic table operations survive optimization, and recovering Luraph-protected scripts with AI-assisted tooling.

nac-l.github.io
2
31
2
151
7,274
Wrote up my attempt at deobfuscating Luraph-protected Lua: lifting Lua to LLVM IR , then using AI + partial evaluation to strip the VM, and checking it against the original. nac-l.github.io/2026/10/12/d…

Deobfuscating Luraph: Lifting Lua to LLVM, Then Letting AI Try

Lifting Lua to LLVM IR, why dynamic table operations survive optimization, and recovering Luraph-protected scripts with AI-assisted tooling.

nac-l.github.io
2
31
2
151
7,274
naci retweeted
Ding
@dingyiOct 11
很期待 omp 作者的新终端，看起来非常牛逼。 stencil.so/tern
44
49
10
614
110,703
naci retweeted
Aarav
@byaaravOct 9
vibe coding final boss
22
319
17
6,298
90,667
naci retweeted
kenopsia@ompuwuOct 9
naci@_nnaciOct 8
Replying to @ompuwu
i will side with claude
1
3
49
naci retweeted
0xSero
@0xSeroOct 9
new awesome meta-harness just dropped stencil.so/tern
35
17
507
38,382
opening omp in tern to deploy shareholder value
6
15
2
193
3,656
naci retweeted
justmazz@justmazzOct 9
jmazzola.github.io/blog/flar… blog post that goes into super detail, should check it out js

a year ago flare-on took 28 hours. this year it took one.

a game dev finished flare-on 13 in about an hour after last year's winner took almost 28, check out my little test that shows how little you need to start reverse engineering well with models now.

jmazzola.github.io
2
5
58
naci retweeted
justmazz@justmazzOct 9
jmazzola.github.io/blog/effo… shameless plug of my blog post that goes into way more detail that twitter will allow me without giving them money for premium

most security is just tedium: my 2027 predictions

a lot of what we call 'secure' is really just 'tedious'. my 2027 predictions on gold farming, emulation, bots, cheats, exploits, n-days, and RSA-1024 once models do the tedious part.

jmazzola.github.io
2
5
45
got a haircut this morning and my barber asked what i do for work. i said ai stuff, agents etc and he goes "oh like a travel agent" and i should have said yeah basically. i did not say yeah basically
2
6
316
more replies
he said his son uses chatgpt for homework and went back to my fade. the 3 guys waiting were arguing about football and never looked up once
1
4
46
anyway came home and spent the 4 hours messing with my omp config so
3
35
Load more