Jack — building in public (@itsjackdev) header
Jack — building in public (@itsjackdev) profile picture

Building AI products in public — code, metrics, and the failures too. First project ships this month.

India
Joined June 2026
Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)Photo by Jack — building in public (@itsjackdev)
Most "AI crypto wallets" are a demo video and a waitlist. INTENT LAYER isn't. I typed 'send 5 USDC to 0x…' into a chat box — it planned the route, signed in my browser, landed on-chain. Real tx, non-custodial. Testnet, three ecosystems, no token. Building it in public.
9
3
17
7,220
Jack — building in public retweeted
keeping his name off-chain
550
234
116
2,016
511,844
Jack — building in public retweeted
if it can't be verified, it doesn't get signed. that's the rule under intent-layer, on testnet. it signs the intent it can verify and refuses the one it can't — "sell if tesla drops 20%" never signs, because it can't price tesla. no override. keys never leave the device.
1
2
160
if it can't be verified, it doesn't get signed. that's the rule under intent-layer, on testnet. it signs the intent it can verify and refuses the one it can't — "sell if tesla drops 20%" never signs, because it can't price tesla. no override. keys never leave the device.
1
2
160
Jack — building in public retweeted
a finance guy joined a video call with his cfo and coworkers and wired out $25m. every face was fake. his was the only real one. we hardened the keys and left the person in the grass — and ai just aimed a factory at the person.
4
1
1
952
a finance guy joined a video call with his cfo and coworkers and wired out $25m. every face was fake. his was the only real one. we hardened the keys and left the person in the grass — and ai just aimed a factory at the person.
4
1
1
952
Jack — building in public retweeted
everyone's fixated on the $8.5m term labs lost. the number i can't shake is 0.5 — half an ETH of voting power. no bug, no phished key. the contracts obeyed the vote. "the DAO voted" needs to stop being a get-out-of-jail card.
2
2
377
everyone's fixated on the $8.5m term labs lost. the number i can't shake is 0.5 — half an ETH of voting power. no bug, no phished key. the contracts obeyed the vote. "the DAO voted" needs to stop being a get-out-of-jail card.
2
2
377
Jack — building in public retweeted
safepal's breach cracked zero wallets. it leaked 39,798 buyers' names and home addresses off a checkout plugin instead. the safe held. somebody just mailed the map. wrote about the merch table nobody in this space guards.
3
5
1,441
safepal's breach cracked zero wallets. it leaked 39,798 buyers' names and home addresses off a checkout plugin instead. the safe held. somebody just mailed the map. wrote about the merch table nobody in this space guards.
3
5
1,441
Jack — building in public retweeted
excited to work together on this. thank you jensen!
343
238
54
5,017
745,309
Jack — building in public retweeted
and right on cue, while i was writing this: ruflo, 67k stars, shipped an mcp bridge that exposed 233 agent tools over http with no auth. one POST, full RCE, every key. the loaded gun wasn't even in a drawer. CVE-2026-59726.
2
2
376
Jack — building in public retweeted
Grok is very good at agentic tasks!
Agentic work is where @grok 4.6 lands hardest, taking the top spot on the Artificial Analysis Agentic Index at 59, tied with Claude Opus 5 Max. The index measures tool use, planning, autonomy and complex problem solving rather than single answers Grok 4.6 completes tasks in ~53 turns and ~0.5bn input tokens on average, against ~103 turns and ~2.0bn for Claude Opus 5 Max Cost of $0.84 per task, putting it on the intelligence versus cost per task Pareto frontier Enterprises buying agents pay per completed task, not per benchmark point. Turn efficiency is what determines whether a long-running workflow is affordable at volume. Two labs now sit at the top of this index with very different cost structures. Buyers get real choice on price for the first time in agentic deployment.
1,158
989
111
8,020
4,001,393
and right on cue, while i was writing this: ruflo, 67k stars, shipped an mcp bridge that exposed 233 agent tools over http with no auth. one POST, full RCE, every key. the loaded gun wasn't even in a drawer. CVE-2026-59726.
2
2
376
Jack — building in public retweeted
defenders can see the future, and have a narrow window to uplevel their cybersecurity practices now. key is to uplevel fundamentals and apply the best AI tools. what we’re doing at OpenAI, and where other organizations can start: blog.gregbrockman.com/the-de…
150
123
40
974
275,643
Jack — building in public retweeted
last tuesday i gave an agent a key that could read every repo we own. not because it needed them. because scoping one is annoying and i was in a hurry. i noticed an hour later. had the revoke page open before i'd finished the thought.
1
3
3
227
last tuesday i gave an agent a key that could read every repo we own. not because it needed them. because scoping one is annoying and i was in a hurry. i noticed an hour later. had the revoke page open before i'd finished the thought.
1
3
3
227
8/9 we keep bracing for the ai that wants to hurt us. that's not the one that gets us. it's the one that just wants to finish the task, holding a key we couldn't be bothered to scope, walking through a door we left open.
1
3
9/9 wrote the whole thing up — the incidents, the real numbers, and what actually holds at t=0: [paste your published article link here]
5