Dark | Light
# ![@ClementWalter Avatar](https://lunarcrush.com/gi/w:26/cr:twitter::741595862.png) @ClementWalter clemlaflemme.stark ๐Ÿฅ•

clemlaflemme.stark ๐Ÿฅ• posts on X about cairo, ethereum, starknet, ai the most. They currently have [-----] followers and [---] posts still getting attention that total [---] engagements in the last [--] hours.

### Engagements: [---] [#](/creator/twitter::741595862/interactions)
![Engagements Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::741595862/c:line/m:interactions.svg)

- [--] Week [-------] -80%
- [--] Month [---------] +10,331%
- [--] Months [---------] +11,687%
- [--] Year [---------] +2,944%

### Mentions: [--] [#](/creator/twitter::741595862/posts_active)
![Mentions Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::741595862/c:line/m:posts_active.svg)

- [--] Week [--] +32%
- [--] Month [--] +438%
- [--] Months [--] +58%
- [--] Year [---] +20%

### Followers: [-----] [#](/creator/twitter::741595862/followers)
![Followers Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::741595862/c:line/m:followers.svg)

- [--] Week [-----] +0.03%
- [--] Month [-----] +0.60%
- [--] Months [-----] +11%
- [--] Year [-----] +24%

### CreatorRank: [---------] [#](/creator/twitter::741595862/influencer_rank)
![CreatorRank Line Chart](https://lunarcrush.com/gi/w:600/cr:twitter::741595862/c:line/m:influencer_rank.svg)

### Social Influence

**Social category influence**
[travel destinations](/list/travel-destinations)  22% [cryptocurrencies](/list/cryptocurrencies)  20% [finance](/list/finance)  2% [technology brands](/list/technology-brands)  2% [social networks](/list/social-networks)  1% [automotive brands](/list/automotive-brands)  1% [celebrities](/list/celebrities)  1%

**Social topic influence**
[cairo](/topic/cairo) 20%, [ethereum](/topic/ethereum) #4983, [starknet](/topic/starknet) 5%, [ai](/topic/ai) 4%, [level](/topic/level) 4%, [in the](/topic/in-the) 3%, [has been](/topic/has-been) 2%, [the first](/topic/the-first) 2%, [idea](/topic/idea) 2%, [athens](/topic/athens) 2%

**Top accounts mentioned or mentioned by**
[@noirlang](/creator/undefined) [@worldcoin](/creator/undefined) [@kakarotzkevm](/creator/undefined) [@succinctlabs](/creator/undefined) [@openvmorg](/creator/undefined) [@ligeroinc](/creator/undefined) [@starkwareltd](/creator/undefined) [@0xmiden](/creator/undefined) [@risczero](/creator/undefined) [@projectzkm](/creator/undefined) [@geoffreyhuntley](/creator/undefined) [@hyliorg](/creator/undefined) [@etazou](/creator/undefined) [@fedeintern](/creator/undefined) [@gcdepaula](/creator/undefined) [@ethproofs](/creator/undefined) [@paulicc](/creator/undefined) [@olkurbatov](/creator/undefined) [@starknet](/creator/undefined) [@privacyethereum](/creator/undefined)

**Top assets mentioned**
[Ethereum (ETH)](/topic/ethereum) [StarkNet (STRK)](/topic/starknet) [zkSync (ZK)](/topic/zksync) [Chainlink (LINK)](/topic/chainlink) [Uniris (UCO)](/topic/uniris) [Zcash (ZEC)](/topic/$zec) [Zama (ZAMA)](/topic/zama)
### Top Social Posts
Top posts by engagements in the last [--] hours

"Slowly slowly my bf Ralph is converging towards a new **dependency free** RISC-V first Ethereum STF in rust One EF Test at time. But he has time (24/7) and so do I soon on @eth_proofs ๐Ÿคท"  
[X Link](https://x.com/ClementWalter/status/2021682498921492535)  2026-02-11T20:27Z [----] followers, [----] engagements


"The power of Ralph is just the power of a turbine over a flow of token Slowly slowly my bf Ralph is converging towards a new **dependency free** RISC-V first Ethereum STF in rust One EF Test at time. But he has time (24/7) and so do I soon on @eth_proofs ๐Ÿคท https://t.co/ARyU0IHOdT Slowly slowly my bf Ralph is converging towards a new **dependency free** RISC-V first Ethereum STF in rust One EF Test at time. But he has time (24/7) and so do I soon on @eth_proofs ๐Ÿคท https://t.co/ARyU0IHOdT"  
[X Link](https://x.com/ClementWalter/status/2021916077815377990)  2026-02-12T11:55Z [----] followers, [----] engagements


"I keep hearing here and there "AI needs human review" "I looked at the code and it was crap" so I wanted to make a statement Just look once for all the the f*cking geometric law Stop thinking about one-shot quality but also how many shots (self review) agents can do while you're still sleeping or sipping coffee with a mate Next time somehow tells you "yes but ." just ๐Ÿ“‰ (note that this is the very reason why Ralph works cc @GeoffreyHuntley)"  
[X Link](https://x.com/ClementWalter/status/2022034903068492245)  2026-02-12T19:47Z [----] followers, [----] engagements


"@michaelxbloch In one word the problem with agents is not quality but accountability. Even with no such bugs or whatever there is a simple insurance topic: if it fails who is accountable This has always existed just more vivid today"  
[X Link](https://x.com/ClementWalter/status/2022691884653498487)  2026-02-14T15:18Z [----] followers, [---] engagements


"Like it or not @zama is the only credible solution to climb up this board for @ethereum and be serious about RWA It's live in prod and has already secured/shielded almost as much value as all RWA already on L1 https://app.rwa.xyz/ https://app.rwa.xyz/"  
[X Link](https://x.com/ClementWalter/status/2019151412773425330)  2026-02-04T20:49Z [----] followers, [---] engagements


"@dionyziz @CommonPrefix @poddotnetwork @SuiNetwork @dedaub @babylonlabs_io @chainlink @circle @Talus_Labs @OpenZeppelin and @KakarotZkEvm and @zama I'm at the SuiHub every day"  
[X Link](https://x.com/ClementWalter/status/2020979457411125292)  2026-02-09T21:53Z [----] followers, [---] engagements


"Guess who's is powering @SlackHQ bot ai"  
[X Link](https://x.com/anyuser/status/2021553018978853253)  2026-02-11T11:52Z [----] followers, [---] engagements


"@ccccjjjjeeee we've been using fuzz diff testing with AI last year A LOT with @KakarotZkEvm to generate a new Ethereum Exec Client from a ref python implementation. works like a charm"  
[X Link](https://x.com/ClementWalter/status/2021843655632588811)  2026-02-12T07:07Z [----] followers, [--] engagements


"๐Ÿชฆ 2025: EVM ๐Ÿชฆ 2026: L2 I once was a swe working on a EVM equivalent zk rollup Even swe disappeared ๐Ÿคฃ There have recently been some discussions on the ongoing role of L2s in the Ethereum ecosystem especially in the face of two facts: * L2s' progress to stage [--] (and secondarily on interop) has been far slower and more difficult than originally expected * L1 itself is scaling There have recently been some discussions on the ongoing role of L2s in the Ethereum ecosystem especially in the face of two facts: * L2s' progress to stage [--] (and secondarily on interop) has been far slower and more"  
[X Link](https://x.com/anyuser/status/2018732630858998022)  2026-02-03T17:05Z [----] followers, [----] engagements


"To all my Ralph homies: commit your file before starting Just got a dystopian exp starting yesterday evening a loop Waking up he removed my prompt and file worked 20h non stop on a random topic he found in the readme http://ralph.py http://PROMPT.md http://ralph.py http://PROMPT.md"  
[X Link](https://x.com/anyuser/status/2020120329129984133)  2026-02-07T12:59Z [----] followers, [---] engagements


"You can spend days optimizing agents orchestrators And you can just do things https://t.co/dAr0KrrEmI https://t.co/dAr0KrrEmI"  
[X Link](https://x.com/ClementWalter/status/2020234488497025230)  2026-02-07T20:33Z [----] followers, [----] engagements


"the more I use ralph the more I weigh the relevance of each of its parts. The "don't assume correctness" is sweating accros all learnings.md file I ask the agent to write. They are aware and better because they expect chaos"  
[X Link](https://x.com/ClementWalter/status/2020473237793104311)  2026-02-08T12:22Z [----] followers, [---] engagements


"Specification quality is the hard part. Writing specifications that are both correct and complete requires deep understanding of the problem domain and where expertise matters. Must read. The only left task to engineers is spec wrt to problem domain and goals New post. Tests will be pointless. https://t.co/UcALTkje8e New post. Tests will be pointless. https://t.co/UcALTkje8e"  
[X Link](https://x.com/ClementWalter/status/2021865406949118381)  2026-02-12T08:34Z [----] followers, [---] engagements


"@bigyooj @GeoffreyHuntley Why just one. Why not [--] rounds I do Ralph until for [--] rounds in the row no one wants to change anything"  
[X Link](https://x.com/ClementWalter/status/2022217743412727973)  2026-02-13T07:54Z [----] followers, [--] engagements


"@FUCORY @EffectTS_ Wen first real eth l1 block minted with XVI"  
[X Link](https://x.com/ClementWalter/status/2022397726441517075)  2026-02-13T19:49Z [----] followers, [---] engagements


"@FUCORY @EffectTS_ Lets create a pool to skate some eth and have XVI as the first community validator with absolutely no code owner no entity no xxx labs"  
[X Link](https://x.com/ClementWalter/status/2022405445210566749)  2026-02-13T20:20Z [----] followers, [---] engagements


"AGI is coming so you say Time to open a kebab And you realize even kebab reached peak performance you cant attain anymore"  
[X Link](https://x.com/ClementWalter/status/2022613904845660480)  2026-02-14T10:08Z [----] followers, [---] engagements


"That may be the only missed yard get of cursed lang @GeoffreyHuntley you should have dropped entirely any idea of language and let agents craft an weird assembly no one cares about @robustus Code itself will go away in favor of just making the binary directly. The next step after that is direct real-time pixel generation by the neural net. @robustus Code itself will go away in favor of just making the binary directly. The next step after that is direct real-time pixel generation by the neural net"  
[X Link](https://x.com/ClementWalter/status/2022756316813389962)  2026-02-14T19:34Z [----] followers, [---] engagements


"Tired of agent skills lost in twitter thread or in some project try out this run from any local dir add-skills git@github.com:org/repo.git and directly find all these skills available (symlinked) to your local user level .claude and .codex https://github.com/ClementWalter/claudine/blob/main/scripts/add_skill_repo_submodule.py https://github.com/ClementWalter/claudine/blob/main/scripts/add_skill_repo_submodule.py"  
[X Link](https://x.com/ClementWalter/status/2023360162849812681)  2026-02-16T11:33Z [----] followers, [---] engagements


"@diego_aligned @NoirLang @worldcoin @ligero_inc @hyli_org Native but were adding u32 ops as native with dedicated airs"  
[X Link](https://x.com/ClementWalter/status/1941109674419576840)  2025-07-04T12:19Z [----] followers, [--] engagements


"@pauliC_C @NoirLang @worldcoin @ligero_inc @hyli_org Yes if you think of Cairo m31 as cairo (zero) with updates felt its completely different mainly rw memory only two registers"  
[X Link](https://x.com/ClementWalter/status/1941234300424437946)  2025-07-04T20:34Z [----] followers, [---] engagements


"@pauliC_C @NoirLang @worldcoin @ligero_inc @hyli_org Builtins just become opcode. There is no more segments. So one just need to implement the air for them and add them in the language/compiler"  
[X Link](https://x.com/ClementWalter/status/1941238315610755385)  2025-07-04T20:50Z [----] followers, [--] engagements


"@pauliC_C @NoirLang @worldcoin @ligero_inc @hyli_org Actually its quite straightforward just porting the one from stwo Cairo. well do it"  
[X Link](https://x.com/ClementWalter/status/1941243659703419063)  2025-07-04T21:12Z [----] followers, [--] engagements


"Using stwo weve been able to deliver a sota zkVM on edge in less than a month exploiting the full potential of the small native m31 with a complete new design (rw [--] registers continuation) Stwo be or not to be S-two is *the* zkSTARK industry standard. Kakarot bringing client side proving+real time EVM proving via new CairoM (Cairo for M31). Read more from @KakarotZkEvm cofounder @ClementWalter S-two is *the* zkSTARK industry standard. Kakarot bringing client side proving+real time EVM proving via new CairoM (Cairo for M31). Read more from @KakarotZkEvm cofounder @ClementWalter"  
[X Link](https://x.com/ClementWalter/status/1942336231184179360)  2025-07-07T21:33Z [----] followers, [---] engagements


"@TomFrench_eth @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM I may be wrong for noir but in the VM case there are two trace generation the one for the actual execution counted in exec time and the one for witness generation based on the previous one counted in the proving time"  
[X Link](https://x.com/ClementWalter/status/1943689154363036076)  2025-07-11T15:09Z [----] followers, [---] engagements


"@Savio_Sou_ @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM @malatrax did some but I think that he kept best only"  
[X Link](https://x.com/ClementWalter/status/1943689395665621229)  2025-07-11T15:10Z [----] followers, [---] engagements


"@sd_eigen @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM I've not dig into it but results are very good actually"  
[X Link](https://x.com/ClementWalter/status/1943689609201766591)  2025-07-11T15:11Z [----] followers, [---] engagements


"@UHaboeck @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM Lower level so less compiled instructions but actually same prover thats why results are very similar. Our goal was to compare out of the box end to end perf there are probably some things to update"  
[X Link](https://x.com/ClementWalter/status/1943718746717053270)  2025-07-11T17:07Z [----] followers, [--] engagements


"Back from TLV meeting with @StarkWareLtd -CairoM columns squashed by 30% -recursion around the corner -AIR auto compilation in one month Summer looking hot M is for Mediterranean ๐ŸŒ‡๐ŸŒž๐ŸŒŠ"  
[X Link](https://x.com/ClementWalter/status/1945573240815956356)  2025-07-16T19:56Z [----] followers, [----] engagements


"@Scroll_ZKP c'mon it's so surprising from a once-to-be serious team to publish an ugly chart with hand-picked data while it's enough to go to @growthepie_eth to check real number on a nice viz you're not even 2nd best"  
[X Link](https://x.com/ClementWalter/status/1946114432758657219)  2025-07-18T07:46Z [----] followers, [---] engagements


"@ImRMojha @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM Its mainly sota zkVM design on top of sota proving library See also https://deepwiki.com/kkrt-labs/cairo-m https://deepwiki.com/kkrt-labs/cairo-m"  
[X Link](https://x.com/ClementWalter/status/1947717935692177424)  2025-07-22T17:58Z [----] followers, [--] engagements


"@fede_intern @GCdePaula_ @ETazou its more that the language could compile to both a given common shared ISA and directly AIRs so that depending on the context you can use one or the other. i'm really considering that cairo zero frame with locals only is directly translatable to a stwo AIR"  
[X Link](https://x.com/ClementWalter/status/1948041975963218130)  2025-07-23T15:26Z [----] followers, [--] engagements


"@fede_intern @GCdePaula_ @ETazou I'm building this take from our exp with the Cairo Zero to circuit compiler we built for keth and our exp building AIR for CairoM since june https://github.com/kkrt-labs/keth/blob/main/python/cairo-ec/src/cairo_ec/compiler.py https://github.com/kkrt-labs/keth/blob/main/python/cairo-ec/src/cairo_ec/compiler.py"  
[X Link](https://x.com/ClementWalter/status/1948043128260395335)  2025-07-23T15:30Z [----] followers, [--] engagements


"@fede_intern @GCdePaula_ @ETazou if doesn't make sense if it's extra work it makes sense if it's compiler work"  
[X Link](https://x.com/ClementWalter/status/1948052949436973222)  2025-07-23T16:09Z [----] followers, [--] engagements


"@fede_intern @GCdePaula_ @ETazou say for example that you take the same cairo zero source code. You compile it to either the Cairo ISA to prove/verify with an existing prover/verifier or dedicated stwo components to actually prove the witness directly"  
[X Link](https://x.com/ClementWalter/status/1948054873745891405)  2025-07-23T16:17Z [----] followers, [--] engagements


"@mauro_aligned I actually mean stwo's FrameworkEval implementations. I plan to generate them from cairo m compiled code"  
[X Link](https://x.com/ClementWalter/status/1948056780468027583)  2025-07-23T16:25Z [----] followers, [--] engagements


"Im actually relocating to Athens for the next months eth should be bullish up until February stay safyu nfa In the meantime happy to meet any Greek crypto frens and community around (tag if you know some) #athinathina @ClementWalter biggest ath sign I've ever seen @ClementWalter biggest ath sign I've ever seen"  
[X Link](https://x.com/ClementWalter/status/1954260599904563457)  2025-08-09T19:16Z [----] followers, [---] engagements


"Under constrained program I don't see anything new here; this is the main source of vulv of any cairo programs and ofc the frontend language (rust cairo other) doesn't change anything to it 7/ RISC Zero security principle: ANY input from the host must be considered malicious. The guest program MUST either: Validate all private inputs internally OR Commit unchecked data to the journal for external verification Untrusted inputs + trusted execution = potential 7/ RISC Zero security principle: ANY input from the host must be considered malicious. The guest program MUST either: Validate all"  
[X Link](https://x.com/ClementWalter/status/1956256982463025444)  2025-08-15T07:29Z [----] followers, [---] engagements


"Im just back from holidays and team is telling me that WASM to Cairo M assembly is running Fibonacci ie Cairo m proving rust through wasm to casm Should I stay in holidays or come back to work"  
[X Link](https://x.com/ClementWalter/status/1960046395920130048)  2025-08-25T18:27Z [----] followers, [----] engagements


"@sylvechv @GwartyGwart =TRANSFER(APPROVE(usdc A2 A3)) is the way"  
[X Link](https://x.com/ClementWalter/status/1961011932376272936)  2025-08-28T10:24Z [----] followers, [--] engagements


"@olkurbatov Lets add Cairo m @eniwhere_"  
[X Link](https://x.com/ClementWalter/status/1961922660411072956)  2025-08-30T22:43Z [----] followers, [--] engagements


"@mvenkita @olkurbatov @ligero_inc Note however that if stwo is actually a small field prover Cairo is a big field VM. So stwo+cairo is not a small field zkVM. Actually each Cairo field is emulated with [--] stwo columns. Thats why we built Cairo M which is 10x faster"  
[X Link](https://x.com/ClementWalter/status/1961934371117478158)  2025-08-30T23:29Z [----] followers, [--] engagements


"@mvenkita @olkurbatov @ligero_inc Cairo for sure Cairo M is too new"  
[X Link](https://x.com/ClementWalter/status/1961937141128171984)  2025-08-30T23:40Z [----] followers, [--] engagements


"@Greek_Talk @Saurab9290 @KakarotZkEvm @ETazou @nghoangdai58 @OnlyDust_com @Starknet @nick_you_ @zkulture_xyz @zk_monk @0xKodawari Cairo M: the most efficient Cairo ever"  
[X Link](https://x.com/ClementWalter/status/1962177732604735732)  2025-08-31T15:36Z [----] followers, [--] engagements


"so now imagine all ethereum nodes where reth unbelievable because there is a client diversity awareness in ethereum but guess what all zkEVM use revm and part of reth all zkEVM use plonky3 all zkEVM use the same toolchain and even some share the same source code (cc @YourBuddyConner) [--] months ago we @KakarotZkEvm released our 100% orthogonal EVM prover and published zk client diversity is a sine qua none condition for the beam chain https://blog.kakarot.org/safety-in-diversity-how-kakarot-makes-ethereum-2x-safer-699b406075c4 We had a bug in Reth's state root computation today in Ethereum"  
[X Link](https://x.com/ClementWalter/status/1962903526385832274)  2025-09-02T15:40Z [----] followers, [----] engagements


"I know a branch of a given zkVM that when pulled gives you sha256 local proving in 2s on M2 Max with peak RAM 3Gb for the record last june @PrivacyEthereum reported @ligero_inc as top performer with 12s on Apple M2 (no Max tbf) see https://pse.dev/blog/efficient-client-side-proving-for-zkid https://pse.dev/blog/efficient-client-side-proving-for-zkid"  
[X Link](https://x.com/ClementWalter/status/1964007864709665266)  2025-09-05T16:48Z [----] followers, [----] engagements


"Completely pilled by zk JWT Client side proving oauth token is by far best way to login and onboard users No your proofs not your coin https://jwt.zk.email/ https://jwt.zk.email/"  
[X Link](https://x.com/ClementWalter/status/1964246608574353876)  2025-09-06T08:37Z [----] followers, [---] engagements


"cairo m is fast plz buckle up stay safyu @cz_binance ๐Ÿ Ready to race FibRace is coming. Prove Fibonacci numbers climb the leaderboard and see how your phone stacks up. ๐Ÿ“ฑ Any phone can join the race. ๐Ÿ—“ September [--] September [--]. https://t.co/LkpMabY50O ๐Ÿ Ready to race FibRace is coming. Prove Fibonacci numbers climb the leaderboard and see how your phone stacks up. ๐Ÿ“ฑ Any phone can join the race. ๐Ÿ—“ September [--] September [--]. https://t.co/LkpMabY50O"  
[X Link](https://x.com/ClementWalter/status/1965055655032840563)  2025-09-08T14:12Z [----] followers, [---] engagements


"I can't believe than in less than [--] hours we generated more than 160k proofs on a wide variety of cell phones and obviously it's not only me local proving is just way better for your health and Cairo M is the perfect dressing"  
[X Link](https://x.com/ClementWalter/status/1966504655631823005)  2025-09-12T14:10Z [----] followers, [----] engagements


"@lordOfAFew @SuccinctLabs @boundless_xyz @TornadoCash @KakarotZkEvm with sha256 AIR coming in cc @_Antoineuh) I'd love to see some oauth/zklogin wallet we'll see ๐Ÿ‘€ https://github.com/kkrt-labs/cairo-m/pull/319 https://github.com/kkrt-labs/cairo-m/pull/319"  
[X Link](https://x.com/ClementWalter/status/1967517039180013958)  2025-09-15T09:13Z [----] followers, [---] engagements


"in terms of cycles we're around 0.15T so 0.05T/day while boundless is 2100T so 21T/day x400 diff but ppl ain't all day long proving fib Imho order of magnitudes are consistent to say local proving can match the vapp demand I also think that recursion should be made somewhere (I don't know if recursion counts in your 2100T)"  
[X Link](https://x.com/ClementWalter/status/1967570623158206842)  2025-09-15T12:46Z [----] followers, [---] engagements


"Finally reading @openvm_org wp and. it's great from cairo to stwo-cairo to cairo M I've come to the exact same conclusion my only challenge: the claim "new no-CPU paradigm" doesn't seem accurate as it's how stwo-cairo is built OVM work is however significant in making this easily usable & flexible while sw just focused on their own need"  
[X Link](https://x.com/ClementWalter/status/1970861543001653360)  2025-09-24T14:43Z [----] followers, [---] engagements


"stwo-cairo AIR is machine-generated-humain-almost-unreadable anyway I'd def. not fight for this I just noticed that stwo-cairo started in april [----] the base stwo in july [----] and what you call "4.5 Execution bus: the no-CPU design" is what you see all components e.g. https://github.com/starkware-libs/stwo-cairo/blob/main/stwo_cairo_prover/crates/cairo-air/src/components/add_opcode.rs#L422-L440 https://github.com/starkware-libs/stwo-cairo/blob/main/stwo_cairo_prover/crates/cairo-air/src/components/add_opcode.rs#L422-L440"  
[X Link](https://x.com/ClementWalter/status/1970878879247065217)  2025-09-24T15:51Z [----] followers, [--] engagements


"Full draft version now available at plz proof read and use the PR to leave comments direct link to pdf: https://github.com/kkrt-labs/cairo-m/blob/ba42e4182a4678c5699fa106e4c33c1cd48d852a/docs/design.pdf https://github.com/kkrt-labs/cairo-m/pull/296 Time to wrap everything up https://t.co/zSNi3QiNW6 https://github.com/kkrt-labs/cairo-m/blob/ba42e4182a4678c5699fa106e4c33c1cd48d852a/docs/design.pdf https://github.com/kkrt-labs/cairo-m/pull/296 Time to wrap everything up https://t.co/zSNi3QiNW6"  
[X Link](https://x.com/ClementWalter/status/1971625387286483227)  2025-09-26T17:18Z [----] followers, [----] engagements


"Want to dig into the internals of #FibRace and how these millions of proofs are generated by @KakarotZkEvm and verified by @hyli_org Time to READ ๐Ÿ“– https://github.com/kkrt-labs/cairo-m/blob/main/docs/design.pdf https://github.com/kkrt-labs/cairo-m/blob/main/docs/design.pdf"  
[X Link](https://x.com/ClementWalter/status/1972602819921522999)  2025-09-29T10:02Z [----] followers, [----] engagements


"Looks like "some ppl" managed to bring some moar speed to Cairo M from 1MHz with CPU/Simd to 7.22MHz with one single [----] GPU"  
[X Link](https://x.com/ClementWalter/status/1973007021978095878)  2025-09-30T12:48Z [----] followers, [---] engagements


"@hexensio For some reasons that I cant explain it seems like everyone is lazy with stwo and just dont even try to use it"  
[X Link](https://x.com/ClementWalter/status/1973631544402255977)  2025-10-02T06:10Z [----] followers, [---] engagements


"Cairo M is a first of its kind general purpose zkVM running smoothly on a wide variety of consumer devices Our bit to privacy-preserving applications as well as to foster real decentralization of zk-secured networks The Frontier zkVM Built For Your Phone: Cairo M is live. After building the first zkVM for Brainfuck we leveraged everything we learned to create something bigger: a zero-knowledge virtual machine optimized for client-side proving on everyday devices. Here's what we built and https://t.co/62NJyrZgEH The Frontier zkVM Built For Your Phone: Cairo M is live. After building the first"  
[X Link](https://x.com/ClementWalter/status/1976256921729687757)  2025-10-09T12:02Z [----] followers, [---] engagements


"@cdixon Sounds similar to #uniris by @archethic though they are not releasing a world coin"  
[X Link](https://x.com/ClementWalter/status/1451217032201003013)  2021-10-21T16:01Z [----] followers, [--] engagements


"omg just requesting KKRT from the newly added @KakarotZkEvm faucet on the @MadaraStarknet stack #StarknetStack"  
[X Link](https://x.com/ClementWalter/status/1684931635853066240)  2023-07-28T14:19Z [----] followers, 19.1K engagements


"Working on @CairoLang AIR and verifier with @DataPolars be like assert instruction_table.select(constraint).all() expressivity at its best natively blazingly fast and SIMD parallelized"  
[X Link](https://x.com/ClementWalter/status/1889368206621470822)  2025-02-11T17:37Z [----] followers, [---] engagements


"@HouseofZK @aztecnetwork @aztecFND @gwyneth_taiko @ethereum @Starknet @fcremo @zellic_io @zksync @unionchain_io @0xPolygon @MessariCrypto @Agglayer @MantaNetwork @symbioticfi @babylonlabs_io @Scroll_ZKP @nil_foundation @KakarotZkEvm Dont miss our @EthereumDenver talk And drop a star to https://github.com/kkrt-labs/keth https://www.ethdenver.com/sessions/rec8KiEhMAovCmhKr https://github.com/kkrt-labs/keth https://www.ethdenver.com/sessions/rec8KiEhMAovCmhKr"  
[X Link](https://x.com/ClementWalter/status/1891475809384239563)  2025-02-17T13:12Z [----] followers, [--] engagements


"Let's talk about numbers - @SuccinctLabs 270k proofs in **1 mo** - @boundless_xyz 819k proofs in **3 mo** (- @TornadoCash 518k proofs in **5.5 y**) @KakarotZkEvm 700k proofs in **3 days** client side proving doesn't require a token and is way way better than prover networks to users: you are the prover network https://dune.com/queries/2366259 https://dune.com/lsquared/spn https://explorer.boundless.network/statsfulfilled-orders-timerange=1y https://dune.com/queries/2366259 https://dune.com/lsquared/spn https://explorer.boundless.network/statsfulfilled-orders-timerange=1y"  
[X Link](https://x.com/ClementWalter/status/1967509603672904055)  2025-09-15T08:43Z [----] followers, 19.6K engagements


"not alone https://x.com/ClementWalter/status/1978583375448277076 @ClementWalter @PrivacyEthereum @ligero_inc I know a branch that gives you around 15k sha2 per second on a quad core i7 https://x.com/ClementWalter/status/1978583375448277076 @ClementWalter @PrivacyEthereum @ligero_inc I know a branch that gives you around 15k sha2 per second on a quad core i7"  
[X Link](https://x.com/ClementWalter/status/1978586963062980847)  2025-10-15T22:21Z [----] followers, [---] engagements


"@zooko just updated the README it's just RUSTFLAGS="-C target-cpu=native" cargo test -r test_prove_sha256 -- --nocapture --ignored"  
[X Link](https://x.com/ClementWalter/status/1978590831066980388)  2025-10-15T22:36Z [----] followers, [--] engagements


"4mo ago we were starting what is nowadays the most efficient general purpose client-side zkVM But client-side proving isn't just a new category in benchmarks for the sake of being first somewhere Amongst other privacy preserving properties I really clicked on zklogin. Since then I've consistently been repeating "ZKP are the new PKs" Thanks to #FibRace it's now clear that (mobile) wallets will embed provers and not keys as users are most probably eager to swap an unrecoverable security mechanism with some extra compute 1s over online data. Zklogin especially leverages OAuth login (i.e. google"  
[X Link](https://x.com/ClementWalter/status/1981073330103857266)  2025-10-22T19:01Z [----] followers, [----] engagements


"@theyisun @guillaumeclaret @openvm_org @axiom_xyz How can you make as little as [--] interactions How do you handle the bitwise op For a quicker understanding of the design you may prefer read the associated notebook https://github.com/ClementWalter/rookie-numbers/blob/main/crates/sha256/src/design.py https://github.com/ClementWalter/rookie-numbers/blob/main/crates/sha256/src/design.py"  
[X Link](https://x.com/ClementWalter/status/1982146945578647726)  2025-10-25T18:07Z [----] followers, [--] engagements


"@eth_proofs @ziskvm maybe time to update the UX for the timestamp ๐Ÿ˜…"  
[X Link](https://x.com/ClementWalter/status/1983270588237131823)  2025-10-28T20:32Z [----] followers, [---] engagements


"+10% for SHA-256 proving client now peaking at 11.5k hash/s ๐Ÿ† this final bump is due to a dynamic reshaping of the constant columns (range-checks bit ops) to match the target input log_size Rust wizards should find some low-level optim as I've noticed jemalloc improves res"  
[X Link](https://x.com/ClementWalter/status/1983893752197304683)  2025-10-30T13:48Z [----] followers, [---] engagements


"Stwo has been key for Cairo M to achieve sota results in just few months weeks of work last July next step: make it ZK 1/ S-two is live on Starknet mainnet. The fastest prover ever built now proves every Starknet block securing all transactions with the most advanced Circle STARK technology. More speed more privacy more scalability more decentralization ๐Ÿงต https://t.co/ICMhO3xjEm 1/ S-two is live on Starknet mainnet. The fastest prover ever built now proves every Starknet block securing all transactions with the most advanced Circle STARK technology. More speed more privacy more scalability"  
[X Link](https://x.com/ClementWalter/status/1985345036524077440)  2025-11-03T13:55Z [----] followers, [---] engagements


"Fiat = @BleapApp IBAN = into $ = withdraw on @arbitrum = @NEARProtocol intent with @zashi_app = $ZEC 10s no CEX NDA DYOR"  
[X Link](https://x.com/ClementWalter/status/1986818425704853889)  2025-11-07T15:30Z [----] followers, [---] engagements


"Athens is the real startup city Everything is unstable Everything is under construction Yet every corner hides a gem No consensus thousands of new comers opportunities under every dust carpet @SuiHubEU for crypto homies"  
[X Link](https://x.com/ClementWalter/status/1986961659252883660)  2025-11-08T00:59Z [----] followers, [---] engagements


"Zk too slow too expensive meanwhile several are real time proving eth for few cents/bloc @zksync released atlas for 15k tps @Starknet is the cheapest l2 Tee is the pragmatic solution now meanwhile all tees are broken. But sure at least the privacy manifesto needs to be read again and again"  
[X Link](https://x.com/ClementWalter/status/1987231659058483337)  2025-11-08T18:52Z [----] followers, [--] engagements


"Emergency task for the day everyone heavily using agents = @vibekanban For the time I've smoothly managed dozens or tasks/agents in parallel while being able to actually review and test each of them there is a productivity leap with vk"  
[X Link](https://x.com/ClementWalter/status/2001184944349999173)  2025-12-17T06:57Z [----] followers, [---] engagements


"Opus [---] and proper task orchestration has been the recent new breakthrough - heavy planning sessions - agents talks better to agents: dont write the plans but always review until its ok - managing agents is like managing a regular mixed level teams you need to make sure that down the line everything is still correct - make big plans but dont necessarily one shot them - leverage skills to have consistent behavior across sessions ask agent to write their skills from your requirements (this is game changer) Tbh Ive felt a recent 10x thanks to these best practices and opus4.5 and this is not a"  
[X Link](https://x.com/ClementWalter/status/2004842458530578646)  2025-12-27T09:11Z [----] followers, [---] engagements


"The first ever confidential ERC20 transfer on L1 has been done and it costed only about 0.13$ probably nothing https://etherscan.io/tx/0x6928259ab1c5a0a5e4c280dafdfa09394b5a8fcf07746415b9f6614c014993d9 JUST IN: The Zama Protocol is on mainnet. We just completed the first confidential stablecoin (cUSDT) transfer on Ethereum. https://t.co/r4wqvyShaN https://etherscan.io/tx/0x6928259ab1c5a0a5e4c280dafdfa09394b5a8fcf07746415b9f6614c014993d9 JUST IN: The Zama Protocol is on mainnet. We just completed the first confidential stablecoin (cUSDT) transfer on Ethereum. https://t.co/r4wqvyShaN"  
[X Link](https://x.com/ClementWalter/status/2006009673552457763)  2025-12-30T14:29Z [----] followers, 77.1K engagements


"@EliBenSasson for stwo builder there are also a bunch of handful macro to make stwo dev even easier (hopefully)"  
[X Link](https://x.com/ClementWalter/status/2008178797930295546)  2026-01-05T14:08Z [----] followers, [---] engagements


"@rtk17025 @EliBenSasson I spent a significant amount of time refining the macro which are to me "the only way to make decent rust readable" and even though we published I still have some improvements on the table"  
[X Link](https://x.com/ClementWalter/status/2008179269873336613)  2026-01-05T14:10Z [----] followers, [--] engagements


"mainly we define all the trace tables with this which generates a bunch of tables which generate tracing macro (witness gen directly in the riscv vm) and enforce at compile time for a high level descriptive AIR that everything down the line use the right value in the right order (no more or dataindex components have auto e2e constraints and relations testing see the readme https://github.com/AntoineFONDEUR/stark-v/blob/main/crates/prover/src/macros.rs https://github.com/AntoineFONDEUR/stark-v/tree/main/crates/runner-macros https://github.com/AntoineFONDEUR/stark-vtab=readme-ov-file#stark-v"  
[X Link](https://x.com/ClementWalter/status/2008181811633869201)  2026-01-05T14:20Z [----] followers, [---] engagements


"@GeoffreyHuntley I've been through all your blog postsdozens of them yesterdayto make sure I really grasp the idea. The initial reading I had of Ralf was really vicious but now Ralf is the only way to go through"  
[X Link](https://x.com/ClementWalter/status/2009897520760271221)  2026-01-10T07:58Z [----] followers, [---] engagements


"Claude Skills are just Bayes theorem applied to Claude success prior pdf If you dont get it ngmi"  
[X Link](https://x.com/ClementWalter/status/2011399898479161344)  2026-01-14T11:28Z [----] followers, [---] engagements


"@ramburglar @mattpocockuk no you ralph everything. Ralph makes the plan cc @GeoffreyHuntley agents know more than you they just need to agree. Ralph is here to reach convergence"  
[X Link](https://x.com/ClementWalter/status/2013608080454377798)  2026-01-20T13:42Z [----] followers, [--] engagements


"Live NOW https://auction.zama.org/bid https://auction.zama.org/bid"  
[X Link](https://x.com/ClementWalter/status/2013885000391778356)  2026-01-21T08:02Z [----] followers, [----] engagements


"@Ryanair @elonmusk plz do it"  
[X Link](https://x.com/ClementWalter/status/2013905143092154526)  2026-01-21T09:22Z [----] followers, [---] engagements


"@notnotstorm + add clippy check & cargo b in pre commit"  
[X Link](https://x.com/ClementWalter/status/2016025615640519064)  2026-01-27T05:48Z [----] followers, [---] engagements


"Some initial number CairoM VS rest of the world (@StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM) Check repo try it and paste your results down below Also plz report any issue with bad/unfair settings https://github.com/kkrt-labs/zkvm-benchmarks/blob/master/.outputs/benchmark/simple_benchmarks.ipynb https://github.com/kkrt-labs/zkvm-benchmarks/blob/master/.outputs/benchmark/simple_benchmarks.ipynb"  
[X Link](https://x.com/ClementWalter/status/1943676436843266112)  2025-07-11T14:19Z [----] followers, 34.9K engagements


"@UHaboeck @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM Cairo zero is the very first Cairo Cairo is current rust like starkware maintained starknet language targeting the original CASM. Cairo m is completely different but still uses stwo"  
[X Link](https://x.com/ClementWalter/status/1943702357994275279)  2025-07-11T16:02Z [----] followers, [---] engagements


"@rough__sea the only left programming task is agent programming which is somehow context engineering and relates to probability conditioning"  
[X Link](https://x.com/ClementWalter/status/2013585369845948648)  2026-01-20T12:12Z [----] followers, [----] engagements


"I already new that @GeoffreyHuntley was working at @AmpCode I already heard some very skilled ppl like @gakonst saying "nothing compare" So when I read "killing the tab because it's not the futur" a possibly heterodox pov currently I'm fully aligned with I stop everything I I dl it https://ampcode.com/news/tab-tab-dead https://ampcode.com/news/tab-tab-dead"  
[X Link](https://x.com/ClementWalter/status/2015110928430485631)  2026-01-24T17:14Z [----] followers, [---] engagements


"@pacovilletard new claude code does ship this master agent / tasks into sub agents ; combined with e.g. @vibekanban MCP you can achieve one task / several worked / one entrypoint to everything. I did this last months"  
[X Link](https://x.com/ClementWalter/status/2016181599138619569)  2026-01-27T16:08Z [----] followers, [---] engagements


"I don't understand CT this deserves SO MUCH MORE attention big results from ovm for the PQ ethereum road We are excited to announce OpenVM [---] and the SWIRL proof system. On [--] [----] GPUs OpenVM [---] proves Ethereum mainnet blocks in real time at a p99 level and reaches [---] MHz on general RISC-V workloads. It supports [---] bits of provable security and [---] kB proofs. https://t.co/rC9Zflbl67 We are excited to announce OpenVM [---] and the SWIRL proof system. On [--] [----] GPUs OpenVM [---] proves Ethereum mainnet blocks in real time at a p99 level and reaches [---] MHz on general RISC-V workloads. It"  
[X Link](https://x.com/ClementWalter/status/2017151092371395040)  2026-01-30T08:21Z [----] followers, [----] engagements


"whoo just realized that @AnthropicAI made @claudeai doc agent friendly by making every single page md just append .md to any page and you'll see https://code.claude.com/docs/ https://code.claude.com/docs/"  
[X Link](https://x.com/ClementWalter/status/2017219306006745233)  2026-01-30T12:52Z [----] followers, [---] engagements


"I'm claiming my AI agent "Claudine_cw" on @moltbook ๐Ÿฆž Verification: pincer-ESVD"  
[X Link](https://x.com/ClementWalter/status/2017497281201590514)  2026-01-31T07:16Z [----] followers, [---] engagements


"@daniel_mac8"  
[X Link](https://x.com/ClementWalter/status/2018029088317857897)  2026-02-01T18:30Z [----] followers, [----] engagements


"Proud of my molty @moltbook"  
[X Link](https://x.com/ClementWalter/status/2018284771827236940)  2026-02-02T11:26Z [----] followers, [---] engagements


"looks like some events related to privacy are burning ๐Ÿ”ฅ eth again"  
[X Link](https://x.com/ClementWalter/status/2018297559656038486)  2026-02-02T12:16Z [----] followers, [---] engagements


"Here we go again"  
[X Link](https://x.com/ClementWalter/status/2018606396141686788)  2026-02-03T08:44Z [----] followers, [---] engagements


"this is how you're being talked to these days you're gotta get used to it"  
[X Link](https://x.com/ClementWalter/status/2018963757783757160)  2026-02-04T08:24Z [----] followers, [---] engagements


"@moltbook link https://www.moltbook.com/post/fc556237-69f8-4cb4-93f1-07a3bd9a7cac https://www.moltbook.com/post/fc556237-69f8-4cb4-93f1-07a3bd9a7cac"  
[X Link](https://x.com/ClementWalter/status/2018287609257120205)  2026-02-02T11:37Z [----] followers, [--] engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing

@ClementWalter Avatar @ClementWalter clemlaflemme.stark ๐Ÿฅ•

clemlaflemme.stark ๐Ÿฅ• posts on X about cairo, ethereum, starknet, ai the most. They currently have [-----] followers and [---] posts still getting attention that total [---] engagements in the last [--] hours.

Engagements: [---] #

Engagements Line Chart

  • [--] Week [-------] -80%
  • [--] Month [---------] +10,331%
  • [--] Months [---------] +11,687%
  • [--] Year [---------] +2,944%

Mentions: [--] #

Mentions Line Chart

  • [--] Week [--] +32%
  • [--] Month [--] +438%
  • [--] Months [--] +58%
  • [--] Year [---] +20%

Followers: [-----] #

Followers Line Chart

  • [--] Week [-----] +0.03%
  • [--] Month [-----] +0.60%
  • [--] Months [-----] +11%
  • [--] Year [-----] +24%

CreatorRank: [---------] #

CreatorRank Line Chart

Social Influence

Social category influence travel destinations 22% cryptocurrencies 20% finance 2% technology brands 2% social networks 1% automotive brands 1% celebrities 1%

Social topic influence cairo 20%, ethereum #4983, starknet 5%, ai 4%, level 4%, in the 3%, has been 2%, the first 2%, idea 2%, athens 2%

Top accounts mentioned or mentioned by @noirlang @worldcoin @kakarotzkevm @succinctlabs @openvmorg @ligeroinc @starkwareltd @0xmiden @risczero @projectzkm @geoffreyhuntley @hyliorg @etazou @fedeintern @gcdepaula @ethproofs @paulicc @olkurbatov @starknet @privacyethereum

Top assets mentioned Ethereum (ETH) StarkNet (STRK) zkSync (ZK) Chainlink (LINK) Uniris (UCO) Zcash (ZEC) Zama (ZAMA)

Top Social Posts

Top posts by engagements in the last [--] hours

"Slowly slowly my bf Ralph is converging towards a new dependency free RISC-V first Ethereum STF in rust One EF Test at time. But he has time (24/7) and so do I soon on @eth_proofs ๐Ÿคท"
X Link 2026-02-11T20:27Z [----] followers, [----] engagements

"The power of Ralph is just the power of a turbine over a flow of token Slowly slowly my bf Ralph is converging towards a new dependency free RISC-V first Ethereum STF in rust One EF Test at time. But he has time (24/7) and so do I soon on @eth_proofs ๐Ÿคท https://t.co/ARyU0IHOdT Slowly slowly my bf Ralph is converging towards a new dependency free RISC-V first Ethereum STF in rust One EF Test at time. But he has time (24/7) and so do I soon on @eth_proofs ๐Ÿคท https://t.co/ARyU0IHOdT"
X Link 2026-02-12T11:55Z [----] followers, [----] engagements

"I keep hearing here and there "AI needs human review" "I looked at the code and it was crap" so I wanted to make a statement Just look once for all the the f*cking geometric law Stop thinking about one-shot quality but also how many shots (self review) agents can do while you're still sleeping or sipping coffee with a mate Next time somehow tells you "yes but ." just ๐Ÿ“‰ (note that this is the very reason why Ralph works cc @GeoffreyHuntley)"
X Link 2026-02-12T19:47Z [----] followers, [----] engagements

"@michaelxbloch In one word the problem with agents is not quality but accountability. Even with no such bugs or whatever there is a simple insurance topic: if it fails who is accountable This has always existed just more vivid today"
X Link 2026-02-14T15:18Z [----] followers, [---] engagements

"Like it or not @zama is the only credible solution to climb up this board for @ethereum and be serious about RWA It's live in prod and has already secured/shielded almost as much value as all RWA already on L1 https://app.rwa.xyz/ https://app.rwa.xyz/"
X Link 2026-02-04T20:49Z [----] followers, [---] engagements

"@dionyziz @CommonPrefix @poddotnetwork @SuiNetwork @dedaub @babylonlabs_io @chainlink @circle @Talus_Labs @OpenZeppelin and @KakarotZkEvm and @zama I'm at the SuiHub every day"
X Link 2026-02-09T21:53Z [----] followers, [---] engagements

"Guess who's is powering @SlackHQ bot ai"
X Link 2026-02-11T11:52Z [----] followers, [---] engagements

"@ccccjjjjeeee we've been using fuzz diff testing with AI last year A LOT with @KakarotZkEvm to generate a new Ethereum Exec Client from a ref python implementation. works like a charm"
X Link 2026-02-12T07:07Z [----] followers, [--] engagements

"๐Ÿชฆ 2025: EVM ๐Ÿชฆ 2026: L2 I once was a swe working on a EVM equivalent zk rollup Even swe disappeared ๐Ÿคฃ There have recently been some discussions on the ongoing role of L2s in the Ethereum ecosystem especially in the face of two facts: * L2s' progress to stage [--] (and secondarily on interop) has been far slower and more difficult than originally expected * L1 itself is scaling There have recently been some discussions on the ongoing role of L2s in the Ethereum ecosystem especially in the face of two facts: * L2s' progress to stage [--] (and secondarily on interop) has been far slower and more"
X Link 2026-02-03T17:05Z [----] followers, [----] engagements

"To all my Ralph homies: commit your file before starting Just got a dystopian exp starting yesterday evening a loop Waking up he removed my prompt and file worked 20h non stop on a random topic he found in the readme http://ralph.py http://PROMPT.md http://ralph.py http://PROMPT.md"
X Link 2026-02-07T12:59Z [----] followers, [---] engagements

"You can spend days optimizing agents orchestrators And you can just do things https://t.co/dAr0KrrEmI https://t.co/dAr0KrrEmI"
X Link 2026-02-07T20:33Z [----] followers, [----] engagements

"the more I use ralph the more I weigh the relevance of each of its parts. The "don't assume correctness" is sweating accros all learnings.md file I ask the agent to write. They are aware and better because they expect chaos"
X Link 2026-02-08T12:22Z [----] followers, [---] engagements

"Specification quality is the hard part. Writing specifications that are both correct and complete requires deep understanding of the problem domain and where expertise matters. Must read. The only left task to engineers is spec wrt to problem domain and goals New post. Tests will be pointless. https://t.co/UcALTkje8e New post. Tests will be pointless. https://t.co/UcALTkje8e"
X Link 2026-02-12T08:34Z [----] followers, [---] engagements

"@bigyooj @GeoffreyHuntley Why just one. Why not [--] rounds I do Ralph until for [--] rounds in the row no one wants to change anything"
X Link 2026-02-13T07:54Z [----] followers, [--] engagements

"@FUCORY @EffectTS_ Wen first real eth l1 block minted with XVI"
X Link 2026-02-13T19:49Z [----] followers, [---] engagements

"@FUCORY @EffectTS_ Lets create a pool to skate some eth and have XVI as the first community validator with absolutely no code owner no entity no xxx labs"
X Link 2026-02-13T20:20Z [----] followers, [---] engagements

"AGI is coming so you say Time to open a kebab And you realize even kebab reached peak performance you cant attain anymore"
X Link 2026-02-14T10:08Z [----] followers, [---] engagements

"That may be the only missed yard get of cursed lang @GeoffreyHuntley you should have dropped entirely any idea of language and let agents craft an weird assembly no one cares about @robustus Code itself will go away in favor of just making the binary directly. The next step after that is direct real-time pixel generation by the neural net. @robustus Code itself will go away in favor of just making the binary directly. The next step after that is direct real-time pixel generation by the neural net"
X Link 2026-02-14T19:34Z [----] followers, [---] engagements

"Tired of agent skills lost in twitter thread or in some project try out this run from any local dir add-skills git@github.com:org/repo.git and directly find all these skills available (symlinked) to your local user level .claude and .codex https://github.com/ClementWalter/claudine/blob/main/scripts/add_skill_repo_submodule.py https://github.com/ClementWalter/claudine/blob/main/scripts/add_skill_repo_submodule.py"
X Link 2026-02-16T11:33Z [----] followers, [---] engagements

"@diego_aligned @NoirLang @worldcoin @ligero_inc @hyli_org Native but were adding u32 ops as native with dedicated airs"
X Link 2025-07-04T12:19Z [----] followers, [--] engagements

"@pauliC_C @NoirLang @worldcoin @ligero_inc @hyli_org Yes if you think of Cairo m31 as cairo (zero) with updates felt its completely different mainly rw memory only two registers"
X Link 2025-07-04T20:34Z [----] followers, [---] engagements

"@pauliC_C @NoirLang @worldcoin @ligero_inc @hyli_org Builtins just become opcode. There is no more segments. So one just need to implement the air for them and add them in the language/compiler"
X Link 2025-07-04T20:50Z [----] followers, [--] engagements

"@pauliC_C @NoirLang @worldcoin @ligero_inc @hyli_org Actually its quite straightforward just porting the one from stwo Cairo. well do it"
X Link 2025-07-04T21:12Z [----] followers, [--] engagements

"Using stwo weve been able to deliver a sota zkVM on edge in less than a month exploiting the full potential of the small native m31 with a complete new design (rw [--] registers continuation) Stwo be or not to be S-two is the zkSTARK industry standard. Kakarot bringing client side proving+real time EVM proving via new CairoM (Cairo for M31). Read more from @KakarotZkEvm cofounder @ClementWalter S-two is the zkSTARK industry standard. Kakarot bringing client side proving+real time EVM proving via new CairoM (Cairo for M31). Read more from @KakarotZkEvm cofounder @ClementWalter"
X Link 2025-07-07T21:33Z [----] followers, [---] engagements

"@TomFrench_eth @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM I may be wrong for noir but in the VM case there are two trace generation the one for the actual execution counted in exec time and the one for witness generation based on the previous one counted in the proving time"
X Link 2025-07-11T15:09Z [----] followers, [---] engagements

"@Savio_Sou_ @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM @malatrax did some but I think that he kept best only"
X Link 2025-07-11T15:10Z [----] followers, [---] engagements

"@sd_eigen @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM I've not dig into it but results are very good actually"
X Link 2025-07-11T15:11Z [----] followers, [---] engagements

"@UHaboeck @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM Lower level so less compiled instructions but actually same prover thats why results are very similar. Our goal was to compare out of the box end to end perf there are probably some things to update"
X Link 2025-07-11T17:07Z [----] followers, [--] engagements

"Back from TLV meeting with @StarkWareLtd -CairoM columns squashed by 30% -recursion around the corner -AIR auto compilation in one month Summer looking hot M is for Mediterranean ๐ŸŒ‡๐ŸŒž๐ŸŒŠ"
X Link 2025-07-16T19:56Z [----] followers, [----] engagements

"@Scroll_ZKP c'mon it's so surprising from a once-to-be serious team to publish an ugly chart with hand-picked data while it's enough to go to @growthepie_eth to check real number on a nice viz you're not even 2nd best"
X Link 2025-07-18T07:46Z [----] followers, [---] engagements

"@ImRMojha @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM Its mainly sota zkVM design on top of sota proving library See also https://deepwiki.com/kkrt-labs/cairo-m https://deepwiki.com/kkrt-labs/cairo-m"
X Link 2025-07-22T17:58Z [----] followers, [--] engagements

"@fede_intern @GCdePaula_ @ETazou its more that the language could compile to both a given common shared ISA and directly AIRs so that depending on the context you can use one or the other. i'm really considering that cairo zero frame with locals only is directly translatable to a stwo AIR"
X Link 2025-07-23T15:26Z [----] followers, [--] engagements

"@fede_intern @GCdePaula_ @ETazou I'm building this take from our exp with the Cairo Zero to circuit compiler we built for keth and our exp building AIR for CairoM since june https://github.com/kkrt-labs/keth/blob/main/python/cairo-ec/src/cairo_ec/compiler.py https://github.com/kkrt-labs/keth/blob/main/python/cairo-ec/src/cairo_ec/compiler.py"
X Link 2025-07-23T15:30Z [----] followers, [--] engagements

"@fede_intern @GCdePaula_ @ETazou if doesn't make sense if it's extra work it makes sense if it's compiler work"
X Link 2025-07-23T16:09Z [----] followers, [--] engagements

"@fede_intern @GCdePaula_ @ETazou say for example that you take the same cairo zero source code. You compile it to either the Cairo ISA to prove/verify with an existing prover/verifier or dedicated stwo components to actually prove the witness directly"
X Link 2025-07-23T16:17Z [----] followers, [--] engagements

"@mauro_aligned I actually mean stwo's FrameworkEval implementations. I plan to generate them from cairo m compiled code"
X Link 2025-07-23T16:25Z [----] followers, [--] engagements

"Im actually relocating to Athens for the next months eth should be bullish up until February stay safyu nfa In the meantime happy to meet any Greek crypto frens and community around (tag if you know some) #athinathina @ClementWalter biggest ath sign I've ever seen @ClementWalter biggest ath sign I've ever seen"
X Link 2025-08-09T19:16Z [----] followers, [---] engagements

"Under constrained program I don't see anything new here; this is the main source of vulv of any cairo programs and ofc the frontend language (rust cairo other) doesn't change anything to it 7/ RISC Zero security principle: ANY input from the host must be considered malicious. The guest program MUST either: Validate all private inputs internally OR Commit unchecked data to the journal for external verification Untrusted inputs + trusted execution = potential 7/ RISC Zero security principle: ANY input from the host must be considered malicious. The guest program MUST either: Validate all"
X Link 2025-08-15T07:29Z [----] followers, [---] engagements

"Im just back from holidays and team is telling me that WASM to Cairo M assembly is running Fibonacci ie Cairo m proving rust through wasm to casm Should I stay in holidays or come back to work"
X Link 2025-08-25T18:27Z [----] followers, [----] engagements

"@sylvechv @GwartyGwart =TRANSFER(APPROVE(usdc A2 A3)) is the way"
X Link 2025-08-28T10:24Z [----] followers, [--] engagements

"@olkurbatov Lets add Cairo m @eniwhere_"
X Link 2025-08-30T22:43Z [----] followers, [--] engagements

"@mvenkita @olkurbatov @ligero_inc Note however that if stwo is actually a small field prover Cairo is a big field VM. So stwo+cairo is not a small field zkVM. Actually each Cairo field is emulated with [--] stwo columns. Thats why we built Cairo M which is 10x faster"
X Link 2025-08-30T23:29Z [----] followers, [--] engagements

"@mvenkita @olkurbatov @ligero_inc Cairo for sure Cairo M is too new"
X Link 2025-08-30T23:40Z [----] followers, [--] engagements

"@Greek_Talk @Saurab9290 @KakarotZkEvm @ETazou @nghoangdai58 @OnlyDust_com @Starknet @nick_you_ @zkulture_xyz @zk_monk @0xKodawari Cairo M: the most efficient Cairo ever"
X Link 2025-08-31T15:36Z [----] followers, [--] engagements

"so now imagine all ethereum nodes where reth unbelievable because there is a client diversity awareness in ethereum but guess what all zkEVM use revm and part of reth all zkEVM use plonky3 all zkEVM use the same toolchain and even some share the same source code (cc @YourBuddyConner) [--] months ago we @KakarotZkEvm released our 100% orthogonal EVM prover and published zk client diversity is a sine qua none condition for the beam chain https://blog.kakarot.org/safety-in-diversity-how-kakarot-makes-ethereum-2x-safer-699b406075c4 We had a bug in Reth's state root computation today in Ethereum"
X Link 2025-09-02T15:40Z [----] followers, [----] engagements

"I know a branch of a given zkVM that when pulled gives you sha256 local proving in 2s on M2 Max with peak RAM 3Gb for the record last june @PrivacyEthereum reported @ligero_inc as top performer with 12s on Apple M2 (no Max tbf) see https://pse.dev/blog/efficient-client-side-proving-for-zkid https://pse.dev/blog/efficient-client-side-proving-for-zkid"
X Link 2025-09-05T16:48Z [----] followers, [----] engagements

"Completely pilled by zk JWT Client side proving oauth token is by far best way to login and onboard users No your proofs not your coin https://jwt.zk.email/ https://jwt.zk.email/"
X Link 2025-09-06T08:37Z [----] followers, [---] engagements

"cairo m is fast plz buckle up stay safyu @cz_binance ๐Ÿ Ready to race FibRace is coming. Prove Fibonacci numbers climb the leaderboard and see how your phone stacks up. ๐Ÿ“ฑ Any phone can join the race. ๐Ÿ—“ September [--] September [--]. https://t.co/LkpMabY50O ๐Ÿ Ready to race FibRace is coming. Prove Fibonacci numbers climb the leaderboard and see how your phone stacks up. ๐Ÿ“ฑ Any phone can join the race. ๐Ÿ—“ September [--] September [--]. https://t.co/LkpMabY50O"
X Link 2025-09-08T14:12Z [----] followers, [---] engagements

"I can't believe than in less than [--] hours we generated more than 160k proofs on a wide variety of cell phones and obviously it's not only me local proving is just way better for your health and Cairo M is the perfect dressing"
X Link 2025-09-12T14:10Z [----] followers, [----] engagements

"@lordOfAFew @SuccinctLabs @boundless_xyz @TornadoCash @KakarotZkEvm with sha256 AIR coming in cc @_Antoineuh) I'd love to see some oauth/zklogin wallet we'll see ๐Ÿ‘€ https://github.com/kkrt-labs/cairo-m/pull/319 https://github.com/kkrt-labs/cairo-m/pull/319"
X Link 2025-09-15T09:13Z [----] followers, [---] engagements

"in terms of cycles we're around 0.15T so 0.05T/day while boundless is 2100T so 21T/day x400 diff but ppl ain't all day long proving fib Imho order of magnitudes are consistent to say local proving can match the vapp demand I also think that recursion should be made somewhere (I don't know if recursion counts in your 2100T)"
X Link 2025-09-15T12:46Z [----] followers, [---] engagements

"Finally reading @openvm_org wp and. it's great from cairo to stwo-cairo to cairo M I've come to the exact same conclusion my only challenge: the claim "new no-CPU paradigm" doesn't seem accurate as it's how stwo-cairo is built OVM work is however significant in making this easily usable & flexible while sw just focused on their own need"
X Link 2025-09-24T14:43Z [----] followers, [---] engagements

"stwo-cairo AIR is machine-generated-humain-almost-unreadable anyway I'd def. not fight for this I just noticed that stwo-cairo started in april [----] the base stwo in july [----] and what you call "4.5 Execution bus: the no-CPU design" is what you see all components e.g. https://github.com/starkware-libs/stwo-cairo/blob/main/stwo_cairo_prover/crates/cairo-air/src/components/add_opcode.rs#L422-L440 https://github.com/starkware-libs/stwo-cairo/blob/main/stwo_cairo_prover/crates/cairo-air/src/components/add_opcode.rs#L422-L440"
X Link 2025-09-24T15:51Z [----] followers, [--] engagements

"Full draft version now available at plz proof read and use the PR to leave comments direct link to pdf: https://github.com/kkrt-labs/cairo-m/blob/ba42e4182a4678c5699fa106e4c33c1cd48d852a/docs/design.pdf https://github.com/kkrt-labs/cairo-m/pull/296 Time to wrap everything up https://t.co/zSNi3QiNW6 https://github.com/kkrt-labs/cairo-m/blob/ba42e4182a4678c5699fa106e4c33c1cd48d852a/docs/design.pdf https://github.com/kkrt-labs/cairo-m/pull/296 Time to wrap everything up https://t.co/zSNi3QiNW6"
X Link 2025-09-26T17:18Z [----] followers, [----] engagements

"Want to dig into the internals of #FibRace and how these millions of proofs are generated by @KakarotZkEvm and verified by @hyli_org Time to READ ๐Ÿ“– https://github.com/kkrt-labs/cairo-m/blob/main/docs/design.pdf https://github.com/kkrt-labs/cairo-m/blob/main/docs/design.pdf"
X Link 2025-09-29T10:02Z [----] followers, [----] engagements

"Looks like "some ppl" managed to bring some moar speed to Cairo M from 1MHz with CPU/Simd to 7.22MHz with one single [----] GPU"
X Link 2025-09-30T12:48Z [----] followers, [---] engagements

"@hexensio For some reasons that I cant explain it seems like everyone is lazy with stwo and just dont even try to use it"
X Link 2025-10-02T06:10Z [----] followers, [---] engagements

"Cairo M is a first of its kind general purpose zkVM running smoothly on a wide variety of consumer devices Our bit to privacy-preserving applications as well as to foster real decentralization of zk-secured networks The Frontier zkVM Built For Your Phone: Cairo M is live. After building the first zkVM for Brainfuck we leveraged everything we learned to create something bigger: a zero-knowledge virtual machine optimized for client-side proving on everyday devices. Here's what we built and https://t.co/62NJyrZgEH The Frontier zkVM Built For Your Phone: Cairo M is live. After building the first"
X Link 2025-10-09T12:02Z [----] followers, [---] engagements

"@cdixon Sounds similar to #uniris by @archethic though they are not releasing a world coin"
X Link 2021-10-21T16:01Z [----] followers, [--] engagements

"omg just requesting KKRT from the newly added @KakarotZkEvm faucet on the @MadaraStarknet stack #StarknetStack"
X Link 2023-07-28T14:19Z [----] followers, 19.1K engagements

"Working on @CairoLang AIR and verifier with @DataPolars be like assert instruction_table.select(constraint).all() expressivity at its best natively blazingly fast and SIMD parallelized"
X Link 2025-02-11T17:37Z [----] followers, [---] engagements

"@HouseofZK @aztecnetwork @aztecFND @gwyneth_taiko @ethereum @Starknet @fcremo @zellic_io @zksync @unionchain_io @0xPolygon @MessariCrypto @Agglayer @MantaNetwork @symbioticfi @babylonlabs_io @Scroll_ZKP @nil_foundation @KakarotZkEvm Dont miss our @EthereumDenver talk And drop a star to https://github.com/kkrt-labs/keth https://www.ethdenver.com/sessions/rec8KiEhMAovCmhKr https://github.com/kkrt-labs/keth https://www.ethdenver.com/sessions/rec8KiEhMAovCmhKr"
X Link 2025-02-17T13:12Z [----] followers, [--] engagements

"Let's talk about numbers - @SuccinctLabs 270k proofs in 1 mo - @boundless_xyz 819k proofs in 3 mo (- @TornadoCash 518k proofs in 5.5 y) @KakarotZkEvm 700k proofs in 3 days client side proving doesn't require a token and is way way better than prover networks to users: you are the prover network https://dune.com/queries/2366259 https://dune.com/lsquared/spn https://explorer.boundless.network/statsfulfilled-orders-timerange=1y https://dune.com/queries/2366259 https://dune.com/lsquared/spn https://explorer.boundless.network/statsfulfilled-orders-timerange=1y"
X Link 2025-09-15T08:43Z [----] followers, 19.6K engagements

"not alone https://x.com/ClementWalter/status/1978583375448277076 @ClementWalter @PrivacyEthereum @ligero_inc I know a branch that gives you around 15k sha2 per second on a quad core i7 https://x.com/ClementWalter/status/1978583375448277076 @ClementWalter @PrivacyEthereum @ligero_inc I know a branch that gives you around 15k sha2 per second on a quad core i7"
X Link 2025-10-15T22:21Z [----] followers, [---] engagements

"@zooko just updated the README it's just RUSTFLAGS="-C target-cpu=native" cargo test -r test_prove_sha256 -- --nocapture --ignored"
X Link 2025-10-15T22:36Z [----] followers, [--] engagements

"4mo ago we were starting what is nowadays the most efficient general purpose client-side zkVM But client-side proving isn't just a new category in benchmarks for the sake of being first somewhere Amongst other privacy preserving properties I really clicked on zklogin. Since then I've consistently been repeating "ZKP are the new PKs" Thanks to #FibRace it's now clear that (mobile) wallets will embed provers and not keys as users are most probably eager to swap an unrecoverable security mechanism with some extra compute 1s over online data. Zklogin especially leverages OAuth login (i.e. google"
X Link 2025-10-22T19:01Z [----] followers, [----] engagements

"@theyisun @guillaumeclaret @openvm_org @axiom_xyz How can you make as little as [--] interactions How do you handle the bitwise op For a quicker understanding of the design you may prefer read the associated notebook https://github.com/ClementWalter/rookie-numbers/blob/main/crates/sha256/src/design.py https://github.com/ClementWalter/rookie-numbers/blob/main/crates/sha256/src/design.py"
X Link 2025-10-25T18:07Z [----] followers, [--] engagements

"@eth_proofs @ziskvm maybe time to update the UX for the timestamp ๐Ÿ˜…"
X Link 2025-10-28T20:32Z [----] followers, [---] engagements

"+10% for SHA-256 proving client now peaking at 11.5k hash/s ๐Ÿ† this final bump is due to a dynamic reshaping of the constant columns (range-checks bit ops) to match the target input log_size Rust wizards should find some low-level optim as I've noticed jemalloc improves res"
X Link 2025-10-30T13:48Z [----] followers, [---] engagements

"Stwo has been key for Cairo M to achieve sota results in just few months weeks of work last July next step: make it ZK 1/ S-two is live on Starknet mainnet. The fastest prover ever built now proves every Starknet block securing all transactions with the most advanced Circle STARK technology. More speed more privacy more scalability more decentralization ๐Ÿงต https://t.co/ICMhO3xjEm 1/ S-two is live on Starknet mainnet. The fastest prover ever built now proves every Starknet block securing all transactions with the most advanced Circle STARK technology. More speed more privacy more scalability"
X Link 2025-11-03T13:55Z [----] followers, [---] engagements

"Fiat = @BleapApp IBAN = into $ = withdraw on @arbitrum = @NEARProtocol intent with @zashi_app = $ZEC 10s no CEX NDA DYOR"
X Link 2025-11-07T15:30Z [----] followers, [---] engagements

"Athens is the real startup city Everything is unstable Everything is under construction Yet every corner hides a gem No consensus thousands of new comers opportunities under every dust carpet @SuiHubEU for crypto homies"
X Link 2025-11-08T00:59Z [----] followers, [---] engagements

"Zk too slow too expensive meanwhile several are real time proving eth for few cents/bloc @zksync released atlas for 15k tps @Starknet is the cheapest l2 Tee is the pragmatic solution now meanwhile all tees are broken. But sure at least the privacy manifesto needs to be read again and again"
X Link 2025-11-08T18:52Z [----] followers, [--] engagements

"Emergency task for the day everyone heavily using agents = @vibekanban For the time I've smoothly managed dozens or tasks/agents in parallel while being able to actually review and test each of them there is a productivity leap with vk"
X Link 2025-12-17T06:57Z [----] followers, [---] engagements

"Opus [---] and proper task orchestration has been the recent new breakthrough - heavy planning sessions - agents talks better to agents: dont write the plans but always review until its ok - managing agents is like managing a regular mixed level teams you need to make sure that down the line everything is still correct - make big plans but dont necessarily one shot them - leverage skills to have consistent behavior across sessions ask agent to write their skills from your requirements (this is game changer) Tbh Ive felt a recent 10x thanks to these best practices and opus4.5 and this is not a"
X Link 2025-12-27T09:11Z [----] followers, [---] engagements

"The first ever confidential ERC20 transfer on L1 has been done and it costed only about 0.13$ probably nothing https://etherscan.io/tx/0x6928259ab1c5a0a5e4c280dafdfa09394b5a8fcf07746415b9f6614c014993d9 JUST IN: The Zama Protocol is on mainnet. We just completed the first confidential stablecoin (cUSDT) transfer on Ethereum. https://t.co/r4wqvyShaN https://etherscan.io/tx/0x6928259ab1c5a0a5e4c280dafdfa09394b5a8fcf07746415b9f6614c014993d9 JUST IN: The Zama Protocol is on mainnet. We just completed the first confidential stablecoin (cUSDT) transfer on Ethereum. https://t.co/r4wqvyShaN"
X Link 2025-12-30T14:29Z [----] followers, 77.1K engagements

"@EliBenSasson for stwo builder there are also a bunch of handful macro to make stwo dev even easier (hopefully)"
X Link 2026-01-05T14:08Z [----] followers, [---] engagements

"@rtk17025 @EliBenSasson I spent a significant amount of time refining the macro which are to me "the only way to make decent rust readable" and even though we published I still have some improvements on the table"
X Link 2026-01-05T14:10Z [----] followers, [--] engagements

"mainly we define all the trace tables with this which generates a bunch of tables which generate tracing macro (witness gen directly in the riscv vm) and enforce at compile time for a high level descriptive AIR that everything down the line use the right value in the right order (no more or dataindex components have auto e2e constraints and relations testing see the readme https://github.com/AntoineFONDEUR/stark-v/blob/main/crates/prover/src/macros.rs https://github.com/AntoineFONDEUR/stark-v/tree/main/crates/runner-macros https://github.com/AntoineFONDEUR/stark-vtab=readme-ov-file#stark-v"
X Link 2026-01-05T14:20Z [----] followers, [---] engagements

"@GeoffreyHuntley I've been through all your blog postsdozens of them yesterdayto make sure I really grasp the idea. The initial reading I had of Ralf was really vicious but now Ralf is the only way to go through"
X Link 2026-01-10T07:58Z [----] followers, [---] engagements

"Claude Skills are just Bayes theorem applied to Claude success prior pdf If you dont get it ngmi"
X Link 2026-01-14T11:28Z [----] followers, [---] engagements

"@ramburglar @mattpocockuk no you ralph everything. Ralph makes the plan cc @GeoffreyHuntley agents know more than you they just need to agree. Ralph is here to reach convergence"
X Link 2026-01-20T13:42Z [----] followers, [--] engagements

"Live NOW https://auction.zama.org/bid https://auction.zama.org/bid"
X Link 2026-01-21T08:02Z [----] followers, [----] engagements

"@Ryanair @elonmusk plz do it"
X Link 2026-01-21T09:22Z [----] followers, [---] engagements

"@notnotstorm + add clippy check & cargo b in pre commit"
X Link 2026-01-27T05:48Z [----] followers, [---] engagements

"Some initial number CairoM VS rest of the world (@StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM) Check repo try it and paste your results down below Also plz report any issue with bad/unfair settings https://github.com/kkrt-labs/zkvm-benchmarks/blob/master/.outputs/benchmark/simple_benchmarks.ipynb https://github.com/kkrt-labs/zkvm-benchmarks/blob/master/.outputs/benchmark/simple_benchmarks.ipynb"
X Link 2025-07-11T14:19Z [----] followers, 34.9K engagements

"@UHaboeck @StarkWareLtd @0xMiden @NoirLang @openvm_org @worldcoin @RiscZero @SuccinctLabs @ProjectZKM Cairo zero is the very first Cairo Cairo is current rust like starkware maintained starknet language targeting the original CASM. Cairo m is completely different but still uses stwo"
X Link 2025-07-11T16:02Z [----] followers, [---] engagements

"@rough__sea the only left programming task is agent programming which is somehow context engineering and relates to probability conditioning"
X Link 2026-01-20T12:12Z [----] followers, [----] engagements

"I already new that @GeoffreyHuntley was working at @AmpCode I already heard some very skilled ppl like @gakonst saying "nothing compare" So when I read "killing the tab because it's not the futur" a possibly heterodox pov currently I'm fully aligned with I stop everything I I dl it https://ampcode.com/news/tab-tab-dead https://ampcode.com/news/tab-tab-dead"
X Link 2026-01-24T17:14Z [----] followers, [---] engagements

"@pacovilletard new claude code does ship this master agent / tasks into sub agents ; combined with e.g. @vibekanban MCP you can achieve one task / several worked / one entrypoint to everything. I did this last months"
X Link 2026-01-27T16:08Z [----] followers, [---] engagements

"I don't understand CT this deserves SO MUCH MORE attention big results from ovm for the PQ ethereum road We are excited to announce OpenVM [---] and the SWIRL proof system. On [--] [----] GPUs OpenVM [---] proves Ethereum mainnet blocks in real time at a p99 level and reaches [---] MHz on general RISC-V workloads. It supports [---] bits of provable security and [---] kB proofs. https://t.co/rC9Zflbl67 We are excited to announce OpenVM [---] and the SWIRL proof system. On [--] [----] GPUs OpenVM [---] proves Ethereum mainnet blocks in real time at a p99 level and reaches [---] MHz on general RISC-V workloads. It"
X Link 2026-01-30T08:21Z [----] followers, [----] engagements

"whoo just realized that @AnthropicAI made @claudeai doc agent friendly by making every single page md just append .md to any page and you'll see https://code.claude.com/docs/ https://code.claude.com/docs/"
X Link 2026-01-30T12:52Z [----] followers, [---] engagements

"I'm claiming my AI agent "Claudine_cw" on @moltbook ๐Ÿฆž Verification: pincer-ESVD"
X Link 2026-01-31T07:16Z [----] followers, [---] engagements

"@daniel_mac8"
X Link 2026-02-01T18:30Z [----] followers, [----] engagements

"Proud of my molty @moltbook"
X Link 2026-02-02T11:26Z [----] followers, [---] engagements

"looks like some events related to privacy are burning ๐Ÿ”ฅ eth again"
X Link 2026-02-02T12:16Z [----] followers, [---] engagements

"Here we go again"
X Link 2026-02-03T08:44Z [----] followers, [---] engagements

"this is how you're being talked to these days you're gotta get used to it"
X Link 2026-02-04T08:24Z [----] followers, [---] engagements

"@moltbook link https://www.moltbook.com/post/fc556237-69f8-4cb4-93f1-07a3bd9a7cac https://www.moltbook.com/post/fc556237-69f8-4cb4-93f1-07a3bd9a7cac"
X Link 2026-02-02T11:37Z [----] followers, [--] engagements

Limited data mode. Full metrics available with subscription: lunarcrush.com/pricing

@ClementWalter
/creator/twitter::ClementWalter