Thats an amazing process you have built in AGNT.
I have recently actually started using AGNT as my harness and I love that it's hitting that RPA feeling with workflow building then adding agents in where needed.
So happy I have been following along since before it was AGNT.
still reverse engineering retro games by hand?
stop wasting months on buggy code, endless debugging, and weird runtime crashes.
introducing automated rom-to-wasm porting by @agnt_gg
slash dev costs from $100k+ down to $1k
compress years of manual labor into 24 hours
hit 95–99.9% native wasm coverage (miles ahead of standard recomps)
just send me over a rom or any executable. no source code needed.
i'll send you back a production ready, native speed wasm recompiler + full emulator runtime within 24 hours.
dm me to port your software to the web.
starts at $1,000 • gba titles around $3k–$5k • n64 ps1 & up $5k+
24h turnaround for most gba and under
if i can't deliver a fully playable build, you pay nothing.
only while supplies last. 50% deposit to get started
example output quality in the video & repo below 👇
@Enscion25 Exactly how I feel.
I haven't seen them change at all.
All I have seen is them double down on the same things they have been saying since the beginning. I never trusted their principles. They are an "ends justify the means" & "we know what's good for you, better than you" corp
NEW: malware developers added nuclear & biological weapons text to to their spyware.
Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner.
Cleanest practical example I can think of for why over-indexing on first order safety alignment is risky.
When closed (and open) models ship with aggressive refusals, they will be sprinkled with second-order blindspots that attackers will discover...and exploit.
We are only in the earliest days of attackers leveraging these features, and it wouldn't surprise me if users systems that need to handle complex cybersecurity issues demand that models be less safety-blunted.
In the weeds: @SocketSecurity's post also shows why intention matters in how you design a malware analysis pipeline to avoid prompt manipulation.
H/T to colleagues that shared this with me https://t.co/f3Aj9TYxU4
Trending now: Agents stay in the loop
Trending next:
- Agents circle back
- Agents collect low-hanging fruits
- Agents touch base
- Agents get on the same page
- Agents reinvent the wheel
@davepl1968@TrpstrLeonOG What do you mean dent them the job?
No one said the person with a roommate or who lives with their parents get denied the job for doing this. On the contrary the same job would afford them a life where they were in that position out of pure choice and not job market pressure.
Lockdown mode is now available in ChatGPT.
We rolled this out for organizations a few months ago, and now it's available for all users on all plans.
Lockdown Mode is designed to help prevent the final stage of data exfiltration from a prompt injection attack by limiting outbound network requests that could transfer sensitive data to an attacker.
Lockdown mode is not meant for everyone. However, for folks who have an elevated risk profile - due to who they are, what they work on, or the types of data they work with - it's an excellent tool for further securing themselves. This has some tradeoffs on functionality and utility, but for these users, the tradeoff is worthwhile.
Proud to see the team ship another wonderful feature as we continue our journey building the most secure, useful, and safest AI capabilities on the planet.
https://t.co/FPOlXI4YoV
@gregisenberg That's just the tip of the iceberg.
Wait until you create entire applications to sort, store, and recal the information and then still never review it.