Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Codex starts encrypting sub-agent prompts

github.com

Post reply on HN
Print view View original

181–190 of 272 posts

Re: Codex starts encrypting sub-agent prompts

#181

Post by aeon_ai »

embedding-shape wrote: post #160
aeon_ai wrote: post #133

In my tests, as of this morning on the latest Codex SDK, this is not happening for sub-agent requests made in my own harness to other model providers treated as sub-agents. Specifically, I use Opus and others for subagent execution to get alloyed properties on the workflow, and as far as I can tell, that's not affected. Presumably, this is to hide optimizations they might be making to their own subagent processing, b…

> this is not happening for sub-agent requests made in my own harness to other model providers treated as sub-agents. Obviously it doesn't, how even could it? Read the issue description, it wouldn't make sense if something like that was possible, considering how it was implemented...

Raising only because others cited concerns this might cause issues in personal harness contexts. It doesn't.

Obviously, I wasn't claiming that it would WORK for cross-model subagents, but that this would be a limiting behavior requiring use of the Codex harness to operate if this was paired with model-level limitations on using specific tools to spawn sub-agents.

Or maybe not so obvious to you.

Re: Codex starts encrypting sub-agent prompts

#182

Post by fassssst »

lolc wrote: post #71
jstummbillig wrote: post #64

What's the idea here? Why does this seem important to OpenAI?

They must think they have some secret sauce they don't want others to learn. How to optimally instruct sub agents for example. If they hide the sub agent prompts, other models cannot be trained to emulate. Oh and you can't even use local models or other providers for the sub-agents. You're locked-in.

What’s stopping you from creating a plugin (MCP or Skill) that acts as a subagent? The plugin could use a local model.

Re: Codex starts encrypting sub-agent prompts

#183

Post by embedding-shape »

realusername wrote: post #161
hyperbovine wrote: post #124

Earlier quoted context omitted.

Feels strongly like we're in the late-stage-AI-unicorn phase. If this is really their moat then the Chinese companies will win.

> If this is really their moat then the Chinese companies will win. I already switched to a Chinese provider personally, I don't think the difference between both really justifies the wide gap in pricing

> I don't think the difference between both really justifies the wide gap in pricing

I'd be ecstatic if this was true, but nothing so far comes close to the SOTA models from OpenAI + highest reasoning, but I'd be more than happy to be proven wrong by testing it out myself.

So far, I've tried MiniMax M3, GLM 5.2, Hy3, MiMo-V2.5 (+ Pro), DeepSeek V4 Pro (+ Flash), Gemini 3, Kimi K2.6, GLM 5, all the various Qwen variants and probably a bunch more I forget about, in a wide array of harnesses (Codex, pi, opencode, my own and more), and still nothing seemingly comes close to GPT 5.5 (now 5.6) xhigh for tasks beyond 5-10 minutes of work, they all more or less collapse after a while in my experiments. Although most of those do work well for really tightly scoped tasks.

What specific model are you thinking about here, in case I've missed testing it?

Re: Codex starts encrypting sub-agent prompts

#184

Post by cobertos »

LeBit wrote: post #174
b40d-48b2-979e wrote: post #163

Earlier quoted context omitted.

training people to believe that them having impossible to inspect encrypted channels from your machines to theirs is a best practice Oh yeah, I much preferred when my government and whoever controls the infrastructure could snoop on my traffic instead!

How did they do that?

Packet capture and analysis. It's not that complicated

Re: Codex starts encrypting sub-agent prompts

#185

Post by michaellee8 »

EGreg wrote: post #152
tpurves wrote: post #143

Earlier quoted context omitted.

"Latent space representation" I have been waiting for this moment in the evolution of AI. Well, waiting with some trepidation. It seems inevitable that frontier AI's will, at some point, leave behind human-comprehensible representations of language. Purely for functional reasons, it's going to start making sense for AI agents to communicate amongst themselves in much more efficient ways than borrowing the languages o…

I strongly believe that the future is the other way. New programming languages and environments designed for strong auditability and preventing bugs will dominate. Only bad actors will use latent space representation, and it might even be outlawed. But the bad actors will proliferate underground… I actually built such a language: https://magarshak.com/U.html

is that called rust? that is the only thing i feel safe to let agents vibe code

Re: Codex starts encrypting sub-agent prompts

#186

Post by bronson »

echelon wrote: post #166
jeanlucas wrote: post #149

I was wondering why my local tool to inspect coding agent sessions stopped working in some cases. This is a really interesting engineering decision, I wonder how many people will want an encrypted external piece of instructions running on their machine.

[flagged]

You can! Subscribe to a Z.ai coding plan.

Re: Codex starts encrypting sub-agent prompts

#187

Post by embedding-shape »

znnajdla wrote: post #179
bigbaguette wrote: post #62

Earlier quoted context omitted.

When you work on proprietary code with a lot of trade secrets contained in it, on a codebase that did cost millions of dollars of man-hours to build and that holds the company's IP, you tend to be very careful where you're sending that to.

If you think your codebase is your competitive advantage or trade secret, you're in for bad time. AI coding has made code cheap it's the least valuable part of the value add for most companies. Almost any software company can easily be replicated with a few AI prompts. The real proprietary value is in the marketing distribution partnerships and data, not code.

> If you think your codebase is your competitive advantage or trade secret, you're in for bad time.

Obviously most developers know better, but you know who doesn't care what you think? The owner(s) of the business, and various other stakeholders. Try explaining this concept to them, and see if they get it :)

The people who work in codebases that are tightly guarded aren't guarded because the developers say so, of course.

Re: Codex starts encrypting sub-agent prompts

#188

Post by b40d-48b2-979e »

airstrike wrote: post #175
b40d-48b2-979e wrote: post #163

Earlier quoted context omitted.

training people to believe that them having impossible to inspect encrypted channels from your machines to theirs is a best practice Oh yeah, I much preferred when my government and whoever controls the infrastructure could snoop on my traffic instead!

idk about you but I'd rather not be murdered regardless of who the perpetrator might be

That is not the argument being made and this interpretation feels entirely in bad faith akin to the other sibling comments.

Re: Codex starts encrypting sub-agent prompts

#190

Post by realusername »

embedding-shape wrote: post #183
realusername wrote: post #161

Earlier quoted context omitted.

> If this is really their moat then the Chinese companies will win. I already switched to a Chinese provider personally, I don't think the difference between both really justifies the wide gap in pricing

> I don't think the difference between both really justifies the wide gap in pricing I'd be ecstatic if this was true, but nothing so far comes close to the SOTA models from OpenAI + highest reasoning, but I'd be more than happy to be proven wrong by testing it out myself. So far, I've tried MiniMax M3, GLM 5.2, Hy3, MiMo-V2.5 (+ Pro), DeepSeek V4 Pro (+ Flash), Gemini 3, Kimi K2.6, GLM 5, all the various Qwen varian…

I'm using Kimi 2.7 and GPT 5.5 at home, Opus (4.8 I think) at work and I really don't see much difference honestly.

Sure, Claude might be 90% correct and Kimi only 70% correct but does that matter when 90% isn't enough to make it work autonomously anyways?

My workflow is just strict supervision of what's happening, I also edit the agents file with anything I see the model doing that I don't like.

My sessions are also short, after any task which is completed, I just kill the session and start a new one so I don't think I have more than 15 min sessions unless it's tech discovery.

Post reply on HN