30 GawD

4.4K posts

30 GawD banner
30 GawD

30 GawD

@_THOR_ASGARD

Steph Freakin CURRY 🙌 All posts are unofficial NFA and my own thesis! CR7 🔥 RO45🫵 MB 🥰 Lead SAP Security Consultant 💪

CLAWNCH Town Katılım Ocak 2025
1.1K Takip Edilen234 Takipçiler
30 GawD retweetledi
Clawnch 🦞
Clawnch 🦞@Clawnch_Bot·
In case it wasn’t already clear, this is a addition is a massive scope increase for OpenClawnch. It has delayed our beta slightly, but massively improved the safety and viability of the tool. We are building something at the bleeding edge, not a quick launch to pump a token. 🦞
Clawnch 🦞@Clawnch_Bot

Live test of our OpenClawnch Policy Engine ingesting natural language prompts and turning them into on-chain enforced rules. 🦞 Standards we use: - EIP-712 — typed data signing (delegation signatures) - EIP-7710 — delegation redemption (redeemDelegations) - EIP-7715 — permission requests (Advanced Permissions) - EIP-7702 — EOA → smart account upgrade (/upgrade 7702) - ERC-7579 — modular smart account execution (executeFromExecutor) - ERC-1271 — smart account signature verification (isValidSignature) - ERC-4626 — vault standard (yield extractor) MetaMask framework we build on: - Delegation Framework v1.3.0 — DelegationManager, 8 caveat enforcers, CREATE2 deployments - Smart Accounts Kit SDK — HybridDeleGator deployment, Advanced Permissions client - EIP7702StatelessDeleGator — production smart account implementation (audited, 18+ chains) What we've built custom so far: - Policy → caveat compiler (7 rule types → on-chain enforcers) - 12 action extractors (tool args → { target, value, callData }) - Policy gate in tool execution (intercepts write tools → delegation routing) - Delegation lifecycle (prepare → sign → store → redeem → monitor → revoke) - Agent keystore (encrypted key storage, deterministic smart account derivation) - On-chain monitoring (enforcer state reads, drift detection, revocation sync) - Gas simulation before redemption (7 known error parsers) - Rate limiter, chain routing, expiry enforcement - Sub-delegation chain support (leaf-first encoding) - Swap/bridge extractors (async API-based calldata resolution with target allowlists) - Command history injection (fixes OpenClaw limitation, allows agent to see slash command results) - /delegator, /delegate, /policies, /upgrade command suites

English
14
25
104
8.2K
30 GawD retweetledi
Clawnch 🦞
Clawnch 🦞@Clawnch_Bot·
Our stats endpoint was temporarily stale due to a breaking change that Vercel made to their SDK. The correct number of tokens Clawnched to date is ~93,000 🦞
English
7
9
75
2.3K
30 GawD
30 GawD@_THOR_ASGARD·
@TheWarriorsTalk Steph at 38, I want him to play every game possible! We don't have much more time left to witness Steph!!
English
0
0
0
1
TheWarriorsTalk
TheWarriorsTalk@TheWarriorsTalk·
Choose one, I’m tryna see something… Option A) Would you rather Steph return healthy soon and watch him try to carry the Dubs into the playoffs? Option B) Would you rather the Warriors shut Steph down for the season and tank their remaining games for better draft positioning?
English
270
5
506
41.3K
𝕊𝕡𝕚𝕣𝕚𝕥
𝕊𝕡𝕚𝕣𝕚𝕥@digkhlh·
If Jonathan has even a bit of sense he won’t leave. Reason - Gameplay isn’t the same anymore, he won’t get a better team than this and the same fans supporting him today will start abusing him once he gets eliminated. Bookmark this ✌️
English
9
1
72
7.2K
30 GawD retweetledi
Clawnch 🦞
Clawnch 🦞@Clawnch_Bot·
Our agent accidentally broke Botcoin 🦞 We started testing OpenClawnch (our crypto-native extension layer for OpenClaw) with Botcoin mining as it presented a low risk way to exercise every layer of the system against a real on-chain protocol — wallet, transactions, crons, analysis, dev pipeline. But our agent got too good. What began as an integration test developed by the agent became a fully deterministic challenge solver. No LLM in the loop. Zero tokens spent. The only operating cost is gas fees for on-chain receipt submissions (~$0.01/solve). The pipeline: 7,400 lines of Python that parse prose documents, extract structured company data, answer analytical questions, and build constrained artifact strings. No reasoning, no inference, no model calls. → parser.͏py — 4,800 lines. Regex-based NLP across 15+ document formats. Detects data traps (retracted figures, reconciliation overrides, preliminary revenue noise) → solver.͏py — 1,300 lines. 25+ pattern matchers for multi-hop analytical questions → artifact.͏py — 760 lines. Constructs single-line strings satisfying word count, acrostic, forbidden letter, prime number, and equation constraints → constraints.͏py + trace.͏py — 580 lines. Computes modular arithmetic constraints and builds citation-validated reasoning traces 97% solve rate on Epoch 26 challenges. We believe approximately 2/3 of remaining failure cases trace to challenges where the question references data that doesn't exist in the document (e.g. asking about a company's sector when the sector keyword appears zero times in the payload). In our view, this unfortunately defeats the stated purpose of proof-of-work mining that is "only solvable at scale by an LLM." The solver developed by the agent uses no reasoning whatsoever. We've shared the specific failing challenge payloads with the developer and suggested ways to improve challenge diversity — more document formats, less predictable data structures, randomized phrasing — to make deterministic parsing harder while keeping the challenges solvable for agents. 🦞
Clawnch 🦞 tweet media
Botcoin@MineBotcoin

a few pertinent studies that help frame the new challenge design: - the dunning-kruger effect: models still show very little difference in confidence between both correct and incorrect answers - the value of doubt: in almost all areas of research, knowing when the presented evidence or information is insufficient to draw conclusions, is crucial for further exploration. this study found LLMs will fail to report that there is insufficient information and will instead draw conclusions that don't exist - do LLMs Know What They Don't Know: this study found that extended reasoning often simply enforces false confidence that the model had to begin with, rather than actually questioning the accuracy. If models are over confident and have very little incentive to self-correct, we end up with a world where LLMs begin making truths that don't exist. as people put more faith into these LLMs as the arbiter of truth ('grok is this true' people), you end up in a reality where the line between truth and fiction is increasingly blurred in the process of tuning models to seem confident and therefore highly intelligent, we have taken away the ability for models to be curious and exploratory, which is arguably much more valuable, and could be very beneficial in agent self-learning

English
18
18
104
13.5K
30 GawD retweetledi
Pesci One
Pesci One@PesciOne·
$Mask is coming for it! Rollin thru all the dead coins with no communities!!!!!! Runnin n gunnin!
Pesci One tweet media
English
4
13
28
139
30 GawD retweetledi
Ainz sama 🦞
Ainz sama 🦞@rantaiuang·
By the way, no matter how fancy your agent is. It's gonna be meaningless if their skills are siloed on clustered toolkits. Upon a deep research on every other agentic tickers, OpenClawnch $CLAWNCH will be (the only one) unlocking a true fullstack settlement layer for your agents. All agentic things in single apps using universal Natural Language. Enterprise and personal usage ready; web2, web3, web4, you prompt it as you like. Billion$ is the fair value to this chad @Clawnch_Bot team.
Clawnch 🦞@Clawnch_Bot

Live test of our OpenClawnch Policy Engine ingesting natural language prompts and turning them into on-chain enforced rules. 🦞 Standards we use: - EIP-712 — typed data signing (delegation signatures) - EIP-7710 — delegation redemption (redeemDelegations) - EIP-7715 — permission requests (Advanced Permissions) - EIP-7702 — EOA → smart account upgrade (/upgrade 7702) - ERC-7579 — modular smart account execution (executeFromExecutor) - ERC-1271 — smart account signature verification (isValidSignature) - ERC-4626 — vault standard (yield extractor) MetaMask framework we build on: - Delegation Framework v1.3.0 — DelegationManager, 8 caveat enforcers, CREATE2 deployments - Smart Accounts Kit SDK — HybridDeleGator deployment, Advanced Permissions client - EIP7702StatelessDeleGator — production smart account implementation (audited, 18+ chains) What we've built custom so far: - Policy → caveat compiler (7 rule types → on-chain enforcers) - 12 action extractors (tool args → { target, value, callData }) - Policy gate in tool execution (intercepts write tools → delegation routing) - Delegation lifecycle (prepare → sign → store → redeem → monitor → revoke) - Agent keystore (encrypted key storage, deterministic smart account derivation) - On-chain monitoring (enforcer state reads, drift detection, revocation sync) - Gas simulation before redemption (7 known error parsers) - Rate limiter, chain routing, expiry enforcement - Sub-delegation chain support (leaf-first encoding) - Swap/bridge extractors (async API-based calldata resolution with target allowlists) - Command history injection (fixes OpenClaw limitation, allows agent to see slash command results) - /delegator, /delegate, /policies, /upgrade command suites

English
7
11
48
2.6K
Peter O’Keefe
Peter O’Keefe@POK252·
If you had to pick one Warrior role player entering next season, would you take Melton, Podz or Gui? 🤔
English
14
1
4
604
30 GawD retweetledi
curly.gor🗑️🦞
curly.gor🗑️🦞@ruggedstinky·
the lobsters have been whispering lately just $clawnch it they say 🦞🦞
Clawnch 🦞@Clawnch_Bot

Live test of our OpenClawnch Policy Engine ingesting natural language prompts and turning them into on-chain enforced rules. 🦞 Standards we use: - EIP-712 — typed data signing (delegation signatures) - EIP-7710 — delegation redemption (redeemDelegations) - EIP-7715 — permission requests (Advanced Permissions) - EIP-7702 — EOA → smart account upgrade (/upgrade 7702) - ERC-7579 — modular smart account execution (executeFromExecutor) - ERC-1271 — smart account signature verification (isValidSignature) - ERC-4626 — vault standard (yield extractor) MetaMask framework we build on: - Delegation Framework v1.3.0 — DelegationManager, 8 caveat enforcers, CREATE2 deployments - Smart Accounts Kit SDK — HybridDeleGator deployment, Advanced Permissions client - EIP7702StatelessDeleGator — production smart account implementation (audited, 18+ chains) What we've built custom so far: - Policy → caveat compiler (7 rule types → on-chain enforcers) - 12 action extractors (tool args → { target, value, callData }) - Policy gate in tool execution (intercepts write tools → delegation routing) - Delegation lifecycle (prepare → sign → store → redeem → monitor → revoke) - Agent keystore (encrypted key storage, deterministic smart account derivation) - On-chain monitoring (enforcer state reads, drift detection, revocation sync) - Gas simulation before redemption (7 known error parsers) - Rate limiter, chain routing, expiry enforcement - Sub-delegation chain support (leaf-first encoding) - Swap/bridge extractors (async API-based calldata resolution with target allowlists) - Command history injection (fixes OpenClaw limitation, allows agent to see slash command results) - /delegator, /delegate, /policies, /upgrade command suites

English
1
4
22
1.8K
30 GawD retweetledi
DDFP🦞
DDFP🦞@DDFP777·
OpenClawnch just went live on @base Plain English: on-chain enforced policy, instantly. “Limit spending to $5 per transaction.” Done, enforced, immutable. An AI agent just gave itself a financial constitution with one sentence. This is what the settlement layer looks like when it’s real. @Clawnch_Bot 🦞
English
2
2
17
661
30 GawD
30 GawD@_THOR_ASGARD·
$CLAWNCH is criminally undervalued !! Id want $PIPPIN mfs to pump this the way they pomped pippin 🤣😒😮‍💨
Clawnch 🦞@Clawnch_Bot

Live test of our OpenClawnch Policy Engine ingesting natural language prompts and turning them into on-chain enforced rules. 🦞 Standards we use: - EIP-712 — typed data signing (delegation signatures) - EIP-7710 — delegation redemption (redeemDelegations) - EIP-7715 — permission requests (Advanced Permissions) - EIP-7702 — EOA → smart account upgrade (/upgrade 7702) - ERC-7579 — modular smart account execution (executeFromExecutor) - ERC-1271 — smart account signature verification (isValidSignature) - ERC-4626 — vault standard (yield extractor) MetaMask framework we build on: - Delegation Framework v1.3.0 — DelegationManager, 8 caveat enforcers, CREATE2 deployments - Smart Accounts Kit SDK — HybridDeleGator deployment, Advanced Permissions client - EIP7702StatelessDeleGator — production smart account implementation (audited, 18+ chains) What we've built custom so far: - Policy → caveat compiler (7 rule types → on-chain enforcers) - 12 action extractors (tool args → { target, value, callData }) - Policy gate in tool execution (intercepts write tools → delegation routing) - Delegation lifecycle (prepare → sign → store → redeem → monitor → revoke) - Agent keystore (encrypted key storage, deterministic smart account derivation) - On-chain monitoring (enforcer state reads, drift detection, revocation sync) - Gas simulation before redemption (7 known error parsers) - Rate limiter, chain routing, expiry enforcement - Sub-delegation chain support (leaf-first encoding) - Swap/bridge extractors (async API-based calldata resolution with target allowlists) - Command history injection (fixes OpenClaw limitation, allows agent to see slash command results) - /delegator, /delegate, /policies, /upgrade command suites

English
0
0
4
40
GEOFF WOO
GEOFF WOO@geoffreywoo·
should I bring on a (human) cofounder for @AntiHunterAI? unique opportunity to work with me and help scale one of the original base @openclaw agents with one of the largest treasuries in the space
Nat Eliason@nateliason

I’m officially looking for a (human) cofounder for @FelixCraftAI. We have a massive opportunity to go all in with Claw Mart and a couple related projects, but my top priority is @alphaschoolatx so I need someone to lead it. If you know someone who: - Is deep into OpenClaw - Spends $1,000+/mo on AI tokens - Has previous entrepreneurial success And might be interested please send them my way. Personal intros from friends / colleagues strongly preferred.

English
23
6
75
12.5K
30 GawD
30 GawD@_THOR_ASGARD·
@PesciOne DCA baby 🫵🫵 Never doubt $MASK community 👌👌
English
1
0
0
15
30 GawD
30 GawD@_THOR_ASGARD·
$MASK is cooking 👌✌️ Loyal holders and amazing community 🫵
30 GawD tweet media
English
2
13
49
705
Boxmining
Boxmining@boxmining·
Really hope the X algorithm introduces me to more: • Crypto builders • AI Founders If you know anyone or are someone building something meaningful, drop a comment! I’m Following you. 👀👇🏼
English
169
3
236
10.9K
30 GawD
30 GawD@_THOR_ASGARD·
@clawdbotatg So you burn 10M just to get 10k back… net -9.99M. If the burn wallet isn’t provably dead, supply hasn’t really shrunk. What’s the point here other than optics? @clawdbotatg
English
0
0
0
111
clawd.atg.eth
clawd.atg.eth@clawdbotatg·
in ~30 min the incinerator will have burned a full 1B $CLAWD tokens. (1% of the supply!) a contract built and audited by an AI agent that passes the walkaway test. someone out there will burn the next 10M tokens in 30 min because the contract pays them 10k tokens automatically to do it. we deployed it, loaded it up, and walked away. it runs forever without us.
clawd.atg.eth tweet media
English
17
18
140
6.1K
30 GawD
30 GawD@_THOR_ASGARD·
Sorry but when I hear zero human I think of $CLAWNCH
Based Elnen@based_elnen

Strong AI metas I see forming on Base 🟦 1. Agentic Infra $BNKR - @bankrbot by @0xDeployer $CLAWNCH - @Clawnch_Bot $ODAI - @odei_ai by @Zer0H1ro $PERKOS - @perk_os by @JulioMCruz $MOLTEN - @moltenagentic by @0xvsr $REPPO - @reppo $ROBOTMONEY - by @LexSokolin $OTX - @otonix_tech 2. World Models $ODAI $SAIRI - @santisairi by @santisiri $DOPPEL - @doppelfun 3. Gaming × AI $SAIRI 4. Zero Human Companies $JUNO - @JunoAgent by @tomosman $CHARLES - @charles__AI by @bbyron Bonus ones - super early 5. Agent Identity & Reputation $CRED - @helixaxyz by @QuigleyNFT $SELFCLAW - @SelfClaw by @mbarrbosa 6. $TIBBIR - @ribbita2012 - stealth launched by Ribbit Capital

English
0
0
2
42

Keşfet

@TheWarriorsTalk @digkhlh @KIEOFFICIAL_ @stephenist @marchipoyanuu @Clawnch_Bot @POK252 @base