Oba

| protocol engineer @zama

[Oba @obatirou]

I am excited to share my first long-form post. This writeup dives into an issue I discovered in several implementations of RFC6979 deterministic ECDSA signatures. Link below.

1/5

[Oba @obatirou]

@PetitPrinceETH Comment tu trouves les positions exactes?
L’abonnement à la newsletter vaut le coup?

[Oba @obatirou]

Super excited for this new chapter!

[Kakarot (Acquired by Zama) @KakarotZkEvm]

https://t.co/4OlNw2gayT

[Oba @obatirou]

@p_misirov Models have a huge influence and doing the study again now with codex/gpt5 would yield different results. Also, how you use it have influence. Context docs + planning is important. It is mostly spec writing. If done correctly the results are great.

[Oba @obatirou]

@real_philogy https://t.co/WYNpbDkBD1

[Tanishk Goyal @0xtanishk]

Ripping this Solidity claude-md for a week now👇
Gives amazing results, specially when guided correctly with good prompts.

Take it out for a spin and give feedback! Also linking some recent PRs, completely driven by claude code.
(Will start including prompts in commits soon)

[Oba @obatirou]

@emo_eth Might still work https://t.co/dV8RxxUpvn
Found a few issues last time I used it

[Oba @obatirou]

The fix on eth-keys was just merged
https://t.co/h36UT6AIEb

[Oba @obatirou]

I am excited to share my first long-form post. This writeup dives into an issue I discovered in several implementations of RFC6979 deterministic ECDSA signatures. Link below.

1/5

[Oba @obatirou]

Sharing my new repo for playing with cryptography. I will be implementing primitives or anything I found interesting there: https://t.co/h18kWntooK
The first one is AES in Solidity from BlockBreakers by @cryptodavidw: https://t.co/fSlmG0Gn2H

[Oba @obatirou]

@0xFlint_ How do you keep track of your time? Any apps/tips?
Genuinely thinking of doing something like this to be more accountable to myself

[Oba @obatirou]

Though for SR, LLMs are wonderful tools to work with for POCs. Needed some guiding but made a POC for diff testing BLS precompiles for Pectra between revm/go-ethereum/python reference
https://t.co/cCuFv30NKX

[Oba @obatirou]

One thing is certain: LLMs help to go faster when writing code and it’s amazing as a dev but they still make mistakes. Some really difficult to spot. Critical systems, hence smart contracts, will still need manual review.

[Oba @obatirou]

Going fast is important but building scalable and secure code is another thing. This needs to be planned from the start and takes time.

[Oba @obatirou]

@BowTiedDevil Really like hypothesis also!
If you want better perf, you could minimize the usage of assume and instead shape your input to the right form (for example d needs to be different from 0) and use composite strategies.

[Oba @obatirou]

Finally made my personal website https://t.co/K9YeuAW4kX

[Oba @obatirou]

Contributed to execution-specs repo recently for a small missing check on y_parity
https://t.co/fPAjYguXSJ

[teddav @0xteddav]

You’ve heard about Binius by @IrreducibleHW but all this talk about “small fields”, “field extensions” and “binary fields” makes no sense? Let me break it down and explain how these concepts accelerate zk-SNARKs 🚀

[Oba @obatirou]

Thank you for the mention @blockthreat
You can read the full newsletter here
https://t.co/cabqxYBolB

[Blockchain Threat Intelligence @blockthreat]

BlockThreat - Week 5, 2025

🪲 Critical bug patched in Geth
📚 Remedy, Curta and other CTF writeups
🤡 Gensler is back teaching at @MIT

https://t.co/z9CYgKcVxK

[Oba @obatirou]

If you want the TDLR go directly to the POC repository https://t.co/MpLmnihEdL.
For the writeup https://t.co/YjJ7NKv1iX.
My post explains the problem, traces the journey, and shares what I learned. Feedback is more than welcome, my DMs are open.

5/5

[Oba @obatirou]

I am excited to share my first long-form post. This writeup dives into an issue I discovered in several implementations of RFC6979 deterministic ECDSA signatures. Link below.

1/5

[Oba @obatirou]

The issue came down to how the nonce k was generated, particularly the influence of message hash as an input to HMAC. noble-curves reduced it mod N (the curve order) but others did not. This leads to different signatures when the input is greater or equal to the curve order.

4/5