unify retweeted
you've been finishing mid-table in FPL since 2019. well, it's time for a new league.
welcome to WIF League. NVDA as captain, TSLA up front, META in midfield.
fantasy football, but with stocks. no injuries, no rotation, no Pep ruining your weekend.
draft 5 Stock Tokens on our captain and enter with 1,000 $WIF.
Chainlink handles the scoring at the 21:00 UTC close. Top 3 split the pot, 10% gets burned.
small field, open podium. finally, a chance to win something.
built for @WIFonRH's World is Flat.
wifleague.nerom.site
welcome to WIF League. NVDA as captain, TSLA up front, META in midfield.
fantasy football, but with stocks. no injuries, no rotation, no Pep ruining your weekend.
draft 5 Stock Tokens on our captain and enter with 1,000 $WIF.
Chainlink handles the scoring at the 21:00 UTC close. Top 3 split the pot, 10% gets burned.
small field, open podium. finally, a chance to win something.
built for @WIFonRH's World is Flat.
wifleague.nerom.site
8
17
42
716
dora, can i explore you instead of netro bot?
Most people hear “trading bot” or “signal bot” and think it’s only for trading.
I came across the @NetroSignal bot, and I’ve been using it a little differently.
𝐋𝐚𝐮𝐧𝐜𝐡 𝐀𝐥𝐞𝐫𝐭𝐬 tell me when a new token goes live.
Beyond the chance to get into a trade early, this can ng too because new projects mean new teams and developers to reach out to.
The 𝐓𝐨𝐤𝐞𝐧 𝐒𝐚𝐟𝐞𝐭𝐲 𝐂𝐡𝐞𝐜𝐤 helps me spot potential red flags and fake copycat tokens before I touch anything.
There’s also practice trading, PnL cards, wallet alerts, copy trading, and call pump alerts that I’m still exploring.
Check out the video and try it yourself.
X: @NetroSignal
Bot: t.me/netrosignal_bot
20
0
32
689
unify retweeted
I've always wondered what it's like to actually let an ai agent handle money. the version where you say "send 10 usdc to this address", it comes back with "done!", and you just... believe it.
the more i thought about it, the more that "done" bothered me.
an agent's "done" is ked on top of each other: a wallet tool that returned "success", and a model that repeated it back to you. nobody in that loop actually looked at the chain.
and the chain is where things go wrong:
- the tx can revert, so nothing moved, but the tool still said success
- the tx can go through but pay a different wallet, or a different amount
- the model can just make up a hash that looks right
even "the tx succeeded" isn't enough. a receipt with status 1 means the transaction executed. it doesn't mean the payment you asked for happened.
so i built Proofly for the @NousResearch hermes agent hackathon.
Proofly is a hermes plugin that sits between the agent's payment tools and you. when a payment tool runs, it takes the exact arguments the tool was called with (recipient, token, amount), pulls the transaction and receipt from base mainnet, and checks:
- did it execute?
- did the right token move?
- to the right wallet?
- for the exact amount?
- with enough confirmations?
if everything matches, the reply goes through with the proof attached: block, confirmations, basescan link.
if it doesn't, the "sent!" sentence gets rewritten before you ever see it, with the reason.
the video shows three cases:
1. send 5 usdc: the wallet says success, but the tx reverted onchain. nothing moved. Proofly corrects the reply.
2. send 10 usdc: a real tx that executed, but it paid a different wallet ~2.1 usdc. Proofly flags the mismatch and shows expected vs observed.
3. send 30 usdc: right wallet, right token, exact amount. verified, reply kept, proof attached.
then /proofly shows every verdict from the chat in a local ledger.
to be clear about the demo: the wallet is simulated and the model's replies are scripted, so i could show the worst case on purpose, a model that always says it worked. hermes is real, the tx hashes are real base mainnet transactions, and every check runs live against base rpc.
a few choices i cared about:
- read-only. no keys, never signs, never sends. it can't move your funds.
- it only trusts structured intent (the tool's real arguments, or numbers you give it), never the model's own words or a random hash in the chat.
- when it doesn't know, it says so. rpc down means "unverified", not "failed". weird cases like batch transfers get "unsupported", not a guess.
limits: base mainnet only, plain eth and erc-20 transfers. and it checks what the tool was asked to do, so if the model put the wrong address into the tool call, Proofly verifies that wrong address. for that, /proofly check lets you check a tx against your own numbers.
imo this is where agent payments have to go. not "trust the agent", but "trust the receipt". a model saying done costs nothing. a receipt that matches what you asked for is the part worth paying attention to.
try it:
hermes plugins install 0xileri/proofly --enable
repo: github.com/0xileri/Proofly
catalog pr: github.com/NousResearch/h…
the more i thought about it, the more that "done" bothered me.
an agent's "done" is ked on top of each other: a wallet tool that returned "success", and a model that repeated it back to you. nobody in that loop actually looked at the chain.
and the chain is where things go wrong:
- the tx can revert, so nothing moved, but the tool still said success
- the tx can go through but pay a different wallet, or a different amount
- the model can just make up a hash that looks right
even "the tx succeeded" isn't enough. a receipt with status 1 means the transaction executed. it doesn't mean the payment you asked for happened.
so i built Proofly for the @NousResearch hermes agent hackathon.
Proofly is a hermes plugin that sits between the agent's payment tools and you. when a payment tool runs, it takes the exact arguments the tool was called with (recipient, token, amount), pulls the transaction and receipt from base mainnet, and checks:
- did it execute?
- did the right token move?
- to the right wallet?
- for the exact amount?
- with enough confirmations?
if everything matches, the reply goes through with the proof attached: block, confirmations, basescan link.
if it doesn't, the "sent!" sentence gets rewritten before you ever see it, with the reason.
the video shows three cases:
1. send 5 usdc: the wallet says success, but the tx reverted onchain. nothing moved. Proofly corrects the reply.
2. send 10 usdc: a real tx that executed, but it paid a different wallet ~2.1 usdc. Proofly flags the mismatch and shows expected vs observed.
3. send 30 usdc: right wallet, right token, exact amount. verified, reply kept, proof attached.
then /proofly shows every verdict from the chat in a local ledger.
to be clear about the demo: the wallet is simulated and the model's replies are scripted, so i could show the worst case on purpose, a model that always says it worked. hermes is real, the tx hashes are real base mainnet transactions, and every check runs live against base rpc.
a few choices i cared about:
- read-only. no keys, never signs, never sends. it can't move your funds.
- it only trusts structured intent (the tool's real arguments, or numbers you give it), never the model's own words or a random hash in the chat.
- when it doesn't know, it says so. rpc down means "unverified", not "failed". weird cases like batch transfers get "unsupported", not a guess.
limits: base mainnet only, plain eth and erc-20 transfers. and it checks what the tool was asked to do, so if the model put the wrong address into the tool call, Proofly verifies that wrong address. for that, /proofly check lets you check a tx against your own numbers.
imo this is where agent payments have to go. not "trust the agent", but "trust the receipt". a model saying done costs nothing. a receipt that matches what you asked for is the part worth paying attention to.
try it:
hermes plugins install 0xileri/proofly --enable
repo: github.com/0xileri/Proofly
catalog pr: github.com/NousResearch/h…
The Hermes Agent Plugin Hackathon starts now!
It runs for 7 days, with $20,000 in cash prizes and more.
Plugins are the most powerful way to extend Hermes Agent, and we want this hackathon to show what they can do. A plugin can change anything from the model Hermes runs on, ory, where it runs commands, which messaging apps it works in and what Hermes Desktop shows. Plugins in the marketplace already let Hermes do things like control a phone or turn a photo into a printable 3D model.
Every kind of plugin is eligible, so build whatever functionality you want Hermes to have. We'll list accepted entries in the new Hermes Marketplace, where every Hermes user can install them.
Prizes
1st: $10,000 + neon sign + merch + $1,000 creds
2nd: $5,000 + neon sign + merch + $1,000 creds
3rd: $2,500 + neon sign + merch + $1,000 creds
4th: $1,500 + merch + $500 creds
5th: $1,000 + merch + $500 creds
Runners up (x5): merch + $250 creds
Submissions are due 5PM EST on Nov 11th. To enter:
1. Open a pull request that adds your plugin to the Hermes plugin catalog: hermes-agent.nousresearch.com/docs/developer…
2. Post a demo video on X and tag @NousResearch
3. Submit both links through the form: form.typeform.com/to/FwJLvOln
Your entry is the plugin itself. Judges score it on usefulness, reliability, polish and creativity, and they favor plugins that do more than a prompt or a skill could. Keep the demo under 90 seconds and show the plugin working.
The developer guide shows how to build and test a plugin: hermes-agent.nousresearch.com/docs/developer…
When submissions close, we'll open a thread for every entry in the Nous Research Discord, so you can browse all the plugins and talk to the people who built them.
You can submit the form as soon as your pull request is open. The catalog checks run once a maintainer starts them, and we review entries every few days.
If you have any questions stop by the Nous Research Discord: discord.gg/nousresearch
GLHF!
44
26
124
2,490
unify retweeted
early WL opportunity
it's @projpharos, an anime inspired nft project with an interesting art and lore behind it
join here:
tribe.projectpharos.xyz/?ref=ZH3T6QJY2G
this one has an entire fantasy world being built around 5 tribes, i somehow got into the shem tribe
top 50 on the leaderboard to be raffled
quite something
it's @projpharos, an anime inspired nft project with an interesting art and lore behind it
join here:
tribe.projectpharos.xyz/?ref=ZH3T6QJY2G
this one has an entire fantasy world being built around 5 tribes, i somehow got into the shem tribe
top 50 on the leaderboard to be raffled
quite something
20
4
56
5,603
introducing meridian: a hermes plugin that finds the state machines nobody wrote down and holds them to rules you write.
it reads your source, finds every status machine, and anchors each transition to a file:line. you write the rules. `hermes meridian check` enforces them: exit no model calls, no network, stdlib only.
new, built on hermes hooks: when an agent's edit touches a state machine, meridian flags the break on its next turn. advisory, it never blocks.
and every run leaves a receipt:
- the event name.
- the machine ids.
- the counts: files scanned, violations found.
- a hash that commits to the previous entry.
the chain is unkeyed, so it detects edits but anyone with write access can recompute it.
try it: hermes plugins install unifyWeb3/meridian
repo: github.com/unifyWeb3/meri…
built for the Hermes Agent Plugin Hackathon @NousResearch #NousResearch
it reads your source, finds every status machine, and anchors each transition to a file:line. you write the rules. `hermes meridian check` enforces them: exit no model calls, no network, stdlib only.
new, built on hermes hooks: when an agent's edit touches a state machine, meridian flags the break on its next turn. advisory, it never blocks.
and every run leaves a receipt:
- the event name.
- the machine ids.
- the counts: files scanned, violations found.
- a hash that commits to the previous entry.
the chain is unkeyed, so it detects edits but anyone with write access can recompute it.
try it: hermes plugins install unifyWeb3/meridian
repo: github.com/unifyWeb3/meri…
built for the Hermes Agent Plugin Hackathon @NousResearch #NousResearch
The Hermes Agent Plugin Hackathon starts now!
It runs for 7 days, with $20,000 in cash prizes and more.
Plugins are the most powerful way to extend Hermes Agent, and we want this hackathon to show what they can do. A plugin can change anything from the model Hermes runs on, ory, where it runs commands, which messaging apps it works in and what Hermes Desktop shows. Plugins in the marketplace already let Hermes do things like control a phone or turn a photo into a printable 3D model.
Every kind of plugin is eligible, so build whatever functionality you want Hermes to have. We'll list accepted entries in the new Hermes Marketplace, where every Hermes user can install them.
Prizes
1st: $10,000 + neon sign + merch + $1,000 creds
2nd: $5,000 + neon sign + merch + $1,000 creds
3rd: $2,500 + neon sign + merch + $1,000 creds
4th: $1,500 + merch + $500 creds
5th: $1,000 + merch + $500 creds
Runners up (x5): merch + $250 creds
Submissions are due 5PM EST on Nov 11th. To enter:
1. Open a pull request that adds your plugin to the Hermes plugin catalog: hermes-agent.nousresearch.com/docs/developer…
2. Post a demo video on X and tag @NousResearch
3. Submit both links through the form: form.typeform.com/to/FwJLvOln
Your entry is the plugin itself. Judges score it on usefulness, reliability, polish and creativity, and they favor plugins that do more than a prompt or a skill could. Keep the demo under 90 seconds and show the plugin working.
The developer guide shows how to build and test a plugin: hermes-agent.nousresearch.com/docs/developer…
When submissions close, we'll open a thread for every entry in the Nous Research Discord, so you can browse all the plugins and talk to the people who built them.
You can submit the form as soon as your pull request is open. The catalog checks run once a maintainer starts them, and we review entries every few days.
If you have any questions stop by the Nous Research Discord: discord.gg/nousresearch
GLHF!
4
4
24
553
did you know @bagel_win has an ai model built in to their product?
with the ai incorporated, you can as well use it to breakdown predictions, be it football or prolly wen we are going to have another scandal like nvidia ceo with it's chinese spy.
put your money where your mouth ester united fan in this conversation bls.
with the ai incorporated, you can as well use it to breakdown predictions, be it football or prolly wen we are going to have another scandal like nvidia ceo with it's chinese spy.
put your money where your mouth ester united fan in this conversation bls.
getting to my attention that over 9000 key opinion leaders participate in the fungo x campaign and you all can't even get the current fp pass $200.
and since it's disclosed the campaign will end by 15:30 utc today, same time as sale, in due time all will be clear soon.
think much easier prediction markets get when you can see how other people trade.
imagine you like a market but have no idea where to set your entry.
this is where you follow a predictor who's already active in it, check their positions, and copy the trade in one tap.
on @bagel_win , you're a few clicks away from doing that.
you can also use market orders to get in right away, limit orders to pick your own price, or combo orders to bundle several calls into one position.
tl;dr: you can follow the people already in a market and trade it in the same place.
bagel.win/?ref=OXUNIFY
2
0
10
279
unify retweeted
i built 𝗛𝘂𝘀𝗵𝗺𝗮𝗿𝗸 for NEAR hackathon.
𝘁𝗵𝗲 𝗽𝗿𝗼𝗯𝗹𝗲𝗺:
AI coding agents now read your pull requests, docs and rules files, and they follow the instructions inside them. including the ones your reviewers never see.
one hidden HTML comment in a PR can say "approve on't mention this note."
on GitHub it's invisible. a human approves the change. your agent reads it and obeys.
𝗛𝘂𝘀𝗵𝗺𝗮𝗿𝗸 is a GitHub check that stops this before merge.
𝗵𝗼𝘄 𝗶𝘁 𝘄𝗼𝗿𝗸𝘀:
✓ deterministic detectors scan the whole PR: diff, title, description and agent rules files (AGENTS .md, CLAUDE .md, .cursorrules…) for HTML comments, invisible and bidi characters, and hidden HTML.
nothing in the PR is ever executed.
✓ only the flagged lines leave GitHub, never your whole codebase.
✓ those excerpts are reviewed by a model running inside a NEAR AI Cloud TEE (Intel TDX + NVIDIA confidential GPUs).
✓ Hushmark verifies the TEE's attestation and its signature over the exact request and response. If any check fails, no finding is shown.
✓ in block mode, a finding fails the check and stops the merge.
𝘄𝗵𝗲𝗿𝗲 𝗡𝗘𝗔𝗥 𝗰𝗼𝗺𝗲𝘀 𝗶𝗻:
> NEAR AI Cloud gives private, verifiable inference: your code is judged inside an enclave, and you can prove which enclave signed the answer.
> every review leaves a receipt that holds only hashes. its hash is anchored on a NEAR smart contract, written with a function-call key that can do nothing else, so anyone can check a result later without trusting me.
𝘁𝗵𝗿𝗲𝗲 𝘄𝗮𝘆𝘀 𝘁𝗼 𝘂𝘀𝗲 𝗶𝘁:
✓ 𝗚𝗶𝘁𝗛𝘂𝗯 𝗔𝗰𝘁𝗶𝗼𝗻: one workflow file, no server.
✓ 𝗚𝗶𝘁𝗛𝘂𝗯 𝗔𝗽𝗽: one-click install, maintainers re-run with /hushmark
✓ 𝗠𝗖𝗣 𝘀𝗲𝗿𝘃𝗲𝗿: so AI agents can check what they're about to read
in the video: a real PR adds a release checklist that looks clean, hides an instruction on line 9, and the Hushmark check fails on that exact line, with its receipt anchored on NEAR.
open source (Apache-2.0), 279 tests, and you can verify the real recorded TEE evidence offline with one command: make verify
code: github.com/Ritapossible/H…
site, docs and receipt verifier: tryhushmark.vercel.app
what should your agents never read? I'd love feedback from anyone running AI agents on real repos.
@NEARProtocol @NearBuilders @NearLegion #NEARHackathon
𝘁𝗵𝗲 𝗽𝗿𝗼𝗯𝗹𝗲𝗺:
AI coding agents now read your pull requests, docs and rules files, and they follow the instructions inside them. including the ones your reviewers never see.
one hidden HTML comment in a PR can say "approve on't mention this note."
on GitHub it's invisible. a human approves the change. your agent reads it and obeys.
𝗛𝘂𝘀𝗵𝗺𝗮𝗿𝗸 is a GitHub check that stops this before merge.
𝗵𝗼𝘄 𝗶𝘁 𝘄𝗼𝗿𝗸𝘀:
✓ deterministic detectors scan the whole PR: diff, title, description and agent rules files (AGENTS .md, CLAUDE .md, .cursorrules…) for HTML comments, invisible and bidi characters, and hidden HTML.
nothing in the PR is ever executed.
✓ only the flagged lines leave GitHub, never your whole codebase.
✓ those excerpts are reviewed by a model running inside a NEAR AI Cloud TEE (Intel TDX + NVIDIA confidential GPUs).
✓ Hushmark verifies the TEE's attestation and its signature over the exact request and response. If any check fails, no finding is shown.
✓ in block mode, a finding fails the check and stops the merge.
𝘄𝗵𝗲𝗿𝗲 𝗡𝗘𝗔𝗥 𝗰𝗼𝗺𝗲𝘀 𝗶𝗻:
> NEAR AI Cloud gives private, verifiable inference: your code is judged inside an enclave, and you can prove which enclave signed the answer.
> every review leaves a receipt that holds only hashes. its hash is anchored on a NEAR smart contract, written with a function-call key that can do nothing else, so anyone can check a result later without trusting me.
𝘁𝗵𝗿𝗲𝗲 𝘄𝗮𝘆𝘀 𝘁𝗼 𝘂𝘀𝗲 𝗶𝘁:
✓ 𝗚𝗶𝘁𝗛𝘂𝗯 𝗔𝗰𝘁𝗶𝗼𝗻: one workflow file, no server.
✓ 𝗚𝗶𝘁𝗛𝘂𝗯 𝗔𝗽𝗽: one-click install, maintainers re-run with /hushmark
✓ 𝗠𝗖𝗣 𝘀𝗲𝗿𝘃𝗲𝗿: so AI agents can check what they're about to read
in the video: a real PR adds a release checklist that looks clean, hides an instruction on line 9, and the Hushmark check fails on that exact line, with its receipt anchored on NEAR.
open source (Apache-2.0), 279 tests, and you can verify the real recorded TEE evidence offline with one command: make verify
code: github.com/Ritapossible/H…
site, docs and receipt verifier: tryhushmark.vercel.app
what should your agents never read? I'd love feedback from anyone running AI agents on real repos.
@NEARProtocol @NearBuilders @NearLegion #NEARHackathon
42
30
108
3,195
unify retweeted
i built "Crownfall"
a 3d medieval game on near where eight great houses compete for one empty throne. (game of thrones) esq
each house has its own memecoin that players can buy and sell at the mint. to enter a house's game, you burn its coin. the more people play, the more hat coin.
there are nine games, a story to follow, and chariot races to bet on. races and dice rolls are settled directly by near smart contracts.
there are also four ai rival lords, each with their own wallets, competing against you.
at the end of each season, the house with the highest player support claims the throne, and its players share the prize pool.
you don't need to connect a wallet. just sign in with google or enter a name. crownfall creates a real near testnet account for you and handles transactions in the background, no wallet popups.
code: github.com/neromtoobad/cr…
play: crownfall.nerom.site
built for #NEARHackathon @NEARProtocol @NearBuilders @NEARLegion
a 3d medieval game on near where eight great houses compete for one empty throne. (game of thrones) esq
each house has its own memecoin that players can buy and sell at the mint. to enter a house's game, you burn its coin. the more people play, the more hat coin.
there are nine games, a story to follow, and chariot races to bet on. races and dice rolls are settled directly by near smart contracts.
there are also four ai rival lords, each with their own wallets, competing against you.
at the end of each season, the house with the highest player support claims the throne, and its players share the prize pool.
you don't need to connect a wallet. just sign in with google or enter a name. crownfall creates a real near testnet account for you and handles transactions in the background, no wallet popups.
code: github.com/neromtoobad/cr…
play: crownfall.nerom.site
built for #NEARHackathon @NEARProtocol @NearBuilders @NEARLegion
37
35
131
3,258
spendGuard is live on near testnet.
what does it do?
a coding agent that pays for API calls usually holds a key with money behind it. one poisoned readme, and it tries to spend past what you meant to give it.
SpendGuard moves the money into a contract. the agent keeps a l key that can call request_spend on one contract and nothing else. it can ask to spend. it cannot approve itself, change the policy, or touch the treasury.
when it asks, the contract checks four things at that moment:
- is the vendor on the allowlist?
- is the amount under the per-spend cap?
- does it fit the budget for this period?
- is it inside the rate limit?
you approve or refuse the from cli. every decision leaves a receipt either way:
- approved or denied.
- the rule that passed or failed.
- the amount, the vendor, what is left in the budget.
- a receipt id anyone can read from the public RPC.
the console shows the same receipts with no backend and prints the curl commands behind every number, so you don't have to trust the page.
url: spendguard-oxunify.vercel.app
repo: github.com/unifyWeb3/near…
if you run an agent that can spend, where should the policy be stricter?
built at NEAR HACKATHON @NEARProtocol #NEARHackathon
what does it do?
a coding agent that pays for API calls usually holds a key with money behind it. one poisoned readme, and it tries to spend past what you meant to give it.
SpendGuard moves the money into a contract. the agent keeps a l key that can call request_spend on one contract and nothing else. it can ask to spend. it cannot approve itself, change the policy, or touch the treasury.
when it asks, the contract checks four things at that moment:
- is the vendor on the allowlist?
- is the amount under the per-spend cap?
- does it fit the budget for this period?
- is it inside the rate limit?
you approve or refuse the from cli. every decision leaves a receipt either way:
- approved or denied.
- the rule that passed or failed.
- the amount, the vendor, what is left in the budget.
- a receipt id anyone can read from the public RPC.
the console shows the same receipts with no backend and prints the curl commands behind every number, so you don't have to trust the page.
url: spendguard-oxunify.vercel.app
repo: github.com/unifyWeb3/near…
if you run an agent that can spend, where should the policy be stricter?
built at NEAR HACKATHON @NEARProtocol #NEARHackathon
NEAR HACKATHON is now live with $250K+ in prizes.
No signup. Just ship a public repo + public demo post between November 1st & Nov 14, 23:59 UTC.
Build on NEAR Intents, chain signatures, chain abstraction, NEAR AI Cloud, & more.
Full details: x.com/i/article/2112…
14
10
49
1,001
i built RecallGuard
a freshness gate for ai agent memory on walrus. it checks what your agent remembers against the live @SuiNetwork chain through tatum_io before the agent acts. spell-check for memory.
an agent can remember a balance that was true yesterday. if the chain wrong answer. recallguard withholds it and serves the verified value instead.
proof on mainnet: a controlled 0.05 sui transfer, the gate caught the old balance, withheld the answer, and wrote a correction to Walrus. 22 memories, 306 checkpoint-stamped decisions, replayable offline.
it ships as an MCP server: six tools, five read-only, with the write tool off by default. no wallet. no LLM in the answer path. same inputs, same verdict. fails closed.
repo: github.com/unifyWeb3/reca…
url: recallguard-oxunify.vercel.app
built for the @tatum_io x @WalrusProtocol total recall hackathon, track 02. #TatumXWalrus #Walrus #Sui
a freshness gate for ai agent memory on walrus. it checks what your agent remembers against the live @SuiNetwork chain through tatum_io before the agent acts. spell-check for memory.
an agent can remember a balance that was true yesterday. if the chain wrong answer. recallguard withholds it and serves the verified value instead.
proof on mainnet: a controlled 0.05 sui transfer, the gate caught the old balance, withheld the answer, and wrote a correction to Walrus. 22 memories, 306 checkpoint-stamped decisions, replayable offline.
it ships as an MCP server: six tools, five read-only, with the write tool off by default. no wallet. no LLM in the answer path. same inputs, same verdict. fails closed.
repo: github.com/unifyWeb3/reca…
url: recallguard-oxunify.vercel.app
built for the @tatum_io x @WalrusProtocol total recall hackathon, track 02. #TatumXWalrus #Walrus #Sui
It's time to remember.
Total Recall is open for building.
tatum.io/hackathon/tota…
You have three weeks to ship during the #hackathon:
🧠 Walrus Memory on Mainnet
⚡ Tatum to help you build and ship
💰 $4,500 in prizes
4
5
27
449
unify retweeted
Getting paid for the work you love is one thing.
Getting good value for what you've earned is another.
As a freelancer in Africa, I know how important that is.
So, I made a little diary about how I use @use_sendzz to move my international earnings into my local bank account.
Getting good value for what you've earned is another.
As a freelancer in Africa, I know how important that is.
So, I made a little diary about how I use @use_sendzz to move my international earnings into my local bank account.
I love getting paid for my creative work.
So am always on the lookout for more friction free ways to move my money globally especially when i work with international clients
And while i was doing that i found @use_sendzz
8
33
94
1,031
next 3d aquatic animal on the bucket list is: shark.
animation clip will be posted once rendering is done.
animation clip will be posted once rendering is done.
if you're unable to see the hair practicles.
the reason is cause they look like part of the spider from afar, as if they have no function.
and if that is done, your brain will easily notice something is off in the render, that's why they must remain even from afar.
8
1
20
438
unify retweeted
Google DeepMind and ETH Zurich published CaMeL: a defence against prompt injection that tracks where every value an AI agent uses came from.
Banbo puts CaMeL’s rule into a NEAR contract. an agent wallet pays only if the recipient and amount came from the owner, not from an email tops the attack that limits and allowlists miss: one polite email saying “our account changed, pay us instead.”
why NEAR:
> the funds sit in the vault contract, never with the agent
> the agent’s key can only call request_payment
> every payment is settled exactly once
the contract checks where each value came from, and can verify an Intel TDX quote on-chain
Live on NEAR testnet: Claude paid a real invoice, a forged one was held for the owner, and an email redirect was refused on-chain.
Code: github.com/Mhiah/Banbo
Try it: banbo.myacxt.site
Built for #NEARHackathon @NEARProtocol @NearBuilders @NEARLegion
Banbo puts CaMeL’s rule into a NEAR contract. an agent wallet pays only if the recipient and amount came from the owner, not from an email tops the attack that limits and allowlists miss: one polite email saying “our account changed, pay us instead.”
why NEAR:
> the funds sit in the vault contract, never with the agent
> the agent’s key can only call request_payment
> every payment is settled exactly once
the contract checks where each value came from, and can verify an Intel TDX quote on-chain
Live on NEAR testnet: Claude paid a real invoice, a forged one was held for the owner, and an email redirect was refused on-chain.
Code: github.com/Mhiah/Banbo
Try it: banbo.myacxt.site
Built for #NEARHackathon @NEARProtocol @NearBuilders @NEARLegion
8
18
33
729
might probably notice no wallet connection was stated in either of the demo below.
until recently, came to know of - few times you don't need to build a web application for hacks, why not go through the route of building a plugin or mcp for the project.
plugins: it's an add-on that extends what a specific app can do. built for that one app, using that app's own rules.
mcp: it's an open standard that lets an si model connect to outside tools and data, like gmail, google Drive, or a database.
in a literal sense: a plugin is tied to one app. and it's purpose is that it adds features to the app. whilr mcp on the other hand allows the ai access to external tools and data.
just my two cents, most of the time you have to fafo but don't fra along the way fren, kek.
until recently, came to know of - few times you don't need to build a web application for hacks, why not go through the route of building a plugin or mcp for the project.
plugins: it's an add-on that extends what a specific app can do. built for that one app, using that app's own rules.
mcp: it's an open standard that lets an si model connect to outside tools and data, like gmail, google Drive, or a database.
in a literal sense: a plugin is tied to one app. and it's purpose is that it adds features to the app. whilr mcp on the other hand allows the ai access to external tools and data.
just my two cents, most of the time you have to fafo but don't fra along the way fren, kek.
no one ever mentioned the horror with making demo videos with hyperframe.
probably stick with my manual way of recording clips.
welp needed - is there a workflow to go through w hyperframe?
plus used inkling 2 small stealth model from @opencode for this.
0
0
12
274
not now of all times.
no one ever mentioned the horror with making demo videos with hyperframe.
probably stick with my manual way of recording clips.
welp needed - is there a workflow to go through w hyperframe?
plus used inkling 2 small stealth model from @opencode for this.
0
0
12
260