Backpressure is all you need

https://www.lucasfcosta.com/blog/backpressure-is-all-you-need

by lucasfcosta • 2 months ago

220 119 2 months ago

Interesting ideas for generalizing goals to reduce human labor in human <—> agent interactions. That said, maybe it is better to set up customized skills and infrastructure for large projects? At our early stage of trying to capture value of agentic systems, the good ideas in this article might be premature optimization.

vermilingua• 2 months ago

It should also reduce the number of low-quality PRs your teammates have to review for details the agent should have caught itself.

Oh boy.

Alifatisk• 2 months ago

Care to elaborate?

atq21192• 2 months ago

That quote shows an utter disregard for basic human decency.

It is the responsibility of the person running the coding agent to make sure the resulting PRs are high quality. Putting that on your team mates, or worse, random open source project maintainers on the internet, is the definition of an extractive contribution.

mpalmer• 2 months ago

They are probably reacting to the laughable idea that by making PRs 20% better (or whatever), devs will continue to review the code with sufficient rigor to catch even the bugs they're supposedly now preventing. Assuming such rigor was ever present in their work!

Put another way, who are they supposed to hire to tell these low quality PRs apart from the high quality ones? Who even knows how to do something like that?!

wellpast• 2 months ago

I’m willing to be wrong but this industry-wide emphasis on AI creative/coding workflows seems way over-engineered.

Ime successful creative execution looks like micro-iterations where each output informs the next creative move.

I can build something incredibly fast from essentially caveman grunt instructions through an LLM harness, iterating as I go.

Optimizing for feeding a huge plan to an agent sounds to me like a net waste of time. And looking over the shoulder of industry peers trying to do this, I don’t see their outputs or throughput some remarkable improvement over what I can produce with minimal fanfare usage.

0x696C69612• 2 months ago

Yeah it's wild watching so many people decide waterfall is great all of a sudden.

NamTaf3• 2 months ago

Never mind stumbling into proper engineering principles like having documented, testable requirements specifications.

bluGill2• 2 months ago

Every large project in the coming back to waterfall. While the problems are certainly known and it was ultimately developed as a straw man, everything else ends up working worse. That said, you shouldn't be thinking pure waterfall as it's drawn up as a strawman, but rather a waterfall variation with feedback loops. But in the end, in very, very many cases, you have to know an end date in order to get things done because so many other things depend on you being done at the same time. If something is going to get done sooner you can't use it anyway without all the other pieces.

fassssst• 2 months ago

The trick is to have 5 of those huge plans running in parallel.

_zoltan_9• 2 months ago

you do not need micro iterations. you can set macro goals and let the agent/LLM/model whatever you want to call it figure it out.

it works.

piazz• 2 months ago

100% agree. Took me ages of working with the agents to circle back around this, which was the best way to get work done before AI automation anyways.

bunderbunder• 2 months ago

I suspect that letting agents spin away unattended for long stretches of time will become less and less popular as more and more companies blow their token budgets and start requiring some answers to difficult questions before agreeing to further loose the purse strings.

artursapek2• 2 months ago

I agree. I have gotten an incredible amount of work done iterating with 5-30 minute long agent tasks. But it requires I stay engaged, and not go chill on the beach, which I guess is a lot of agentmaxxers’ goal.

mewpmewp2• 2 months ago

For me it's usually that I start with a single agent, but then I won't have anything to do while it is churning and I have other ideas/features that keep building up that I want to do, so I need to scale, and while I'm scaling I need to start to have those workflows, so eventually I end up with many agents, most which are autonomous working on their own worktrees, but I will have a specific agent that I will talk to more iteratively.

So e.g. I may have 1 agent that I ask and iterate on with directly, and 9 agents that work separately on their own.

I will utilize this 1 agent on features I care most about and want to guide and iterate on in as much detail as possible.

Yokohiii2• 2 months ago

LLMs are too flaky for high quality code. On tougher problems it's very common for an LLM to contradict itself and run in circles. It simply doesn't know what the right thing is, but on each turn it is super confident to do the right thing.

Maybe I've chosen hardmode to learn C with LLM assistance, plus my pet project turned out to be a bit less trivial then anticipated. But I know that I have to think three times about my choices how to deal with C problems and seeing how a LLM struggles to give reasonable answers is a a huge red flag and forces me to think about it a fourth time.

Doing all this with a fast autonomous workflow with just little user guidance is asking for trouble.

coffeefirst• 2 months ago

It’s not just you. My last dumb pilot program making a Pocket clone in Python also got stuck in a loop regularly, which should be its strong suit...

I suspect that the “right” way to use LMs in coding, including accounting for focus, control, and costs is not a settled debate. We probably haven’t even seen the best ideas yet. But I’m really dislike the maximalist approach.

doom2• 2 months ago

Maybe I've chosen hardmode to learn C with LLM assistance

May you speak a little more about how you're approaching this? I was thinking of doing similar

dnnddidiej2• 2 months ago

Oh this is 101. Anyone not doing this? If not do it now!

Arodex• 2 months ago

Because your token use explodes?

dnnddidiej2• 2 months ago

Can be done on a claude pro. But if you are low on tokens then yeah probably stick to more of a non-thinking copilot type arrangement (which is fine!)

_zoltan_9• 2 months ago

"In this post, I’ll cover a third, not-so-obvious approach: building ways for the agent to validate more of its own work before a human has to step in. "

this has been an obvious thing to do since at least January (since Geoffrey Huntley published "everything is a ralph loop"), and this is how I've been working: build enough orchestration tooling to be able to automate everything: development container bringup, building it, running the unit tests, doing integration testing, and using the software as eventually an end user. then to iterate set performance goals on an already solid basis so the automated agent ("gym") can go and iterate autonomously, and let you know when it's "done".

I understand this probably does not work if you're on some subscription and not using the API (tokens burn fast), but this has been extremely productive for me.

psychoslave• 2 months ago

What license do you use then?

_zoltan_9• 2 months ago

you can pay by just volume ("API pricing")

kami233• 2 months ago

This is where most of my productivity gains have come, I have a special harness I move from project to project now that does my testing orchestration, lots of my work day is setting up a prompt or two early and just letting them loop till they return evidence that the feature is working having gone through the big QA loop.

I've slowly been optimizing for token use through the stack and Claude ends up making very tight for loops for most of the process and keeping token count even lower. It's been nice. A lot of my toil at work is just gone.

osigurdson4• 2 months ago

I can see how you could avoid regressions this way, but what do you add to your harness to prove that a new feature is working?

dirtbag__dad• 2 months ago

You can get really far with the 20x Claude Code and Codex plans. They are many orders of magnitude cheaper than api calls.

_zoltan_9• 2 months ago

Agreed! Until they fit. :-)

oblio2• 2 months ago

Enjoy that until the token economy comes crashing down on you.

lovich3• 2 months ago

What’s the cost of all that though? I don’t doubt that productivity could be gained but when I see articles like the one on the Open Claw guy spending 1.3 million on tokens in a single month I am reminded of drag racing engines that can reach incredible speeds but also need to be completely rebuilt after a single race.

ip26• 2 months ago

Depends on the quality of your validation loop. Can the agent find the bug in a five second unit test, or does it have to run the full deployment test?

It also presents tradeoffs in compute budget. Cycles spent executing large arrays of tests could mean less tokens spent debugging.

_zoltan_9• 2 months ago

I can't give exact numbers but I find the cost more than OK.

edfletcher_t137• 2 months ago

yep, this has been obvious to a lot of people for awhile. especially after Cherny posted about exactly this in a massively-popular thread... four months ago: https://x.com/bcherny/status/2007179861115511237

dodu_• 2 months ago

I understand this probably does not work if you're on some subscription and not using the API (tokens burn fast), but this has been extremely productive for me.

Do you have infinite money?

_zoltan_9• 2 months ago

compared to salaries, our bill isn't outrageous.

teleforce• 2 months ago

The rest of the paragraph explanations are more important.

"The goal to make longer unattended sessions safe enough to be useful without fully removing the human from the loop. It should also reduce the number of low-quality PRs your teammates have to review for details the agent should have caught itself."

safe enough to be useful without fully removing the human from the loop

This is the fundamental concept for AI usage, assistance and adoption for every fields not only code generation.

Essentially AI including LLM, ML, DL, is just a tool, like any other automation tools operating based on the principle of expert-in-the-loop as safety and quality gatekeeper, for sensible and responsible decision making [1].

[1] Domain expertise has always been the real moat (brethorsting.com) (519 comments):

https://news.ycombinator.com/item?id=48340411

_zoltan_9• 2 months ago

it's fine to remove the human from the loop. set a macro goal, tell the agent how you think it could go there, and let it go nuts.

with enough scaffolding around self-reflectivity and metrics, it will converge.

closeparen• 2 months ago

It is too bad that so many companies have built these huge microservices beasts where you cannot run much more than unit tests locally, and you pretty much have to guess at the impact of your change until you have merged it, deployed it, and turned on the feature flag for your own account / test account. This loop is slow, which is perhaps a cost they are willing to pay, but it is also not safe for agents, which will be a massive setback.

denysvitali• 2 months ago

This seems to be the coding agents 101: build a strong feedback loop. Am I missing something?

artursapek2• 2 months ago

Yeah I don’t really see the backpressure analogy here - it implies that the agent is constantly producing new stuff, which isn’t really possible since the solution is very detailed specs/goals.

lucamark• 2 months ago

No, there is nothing new in this article. This methodology is already adopted - and further optimized - by the sw community.

cyanydeez• 2 months ago

interesting idea, unfortunately programming the structure is equivalent (P=NP) to just programming itself. same as TDD.

as usual, the tool isnt really doing whats listed on its label.

however, people are different so this might improve someones capability to deploy LLMs. might even provide better evidence where actual brain power is needed.

xg15• 2 months ago

Isn't this a bit of an incorrect usage of the term "backpressure"?

OP quoted the correct definition right at the start:

In systems engineering, backpressure is the mechanism by which a downstream component signals upstream that it can't accept more work

(the "downstream component" being the human reviewer in this case)

But the measures they propose don't actually do that. They are more like fixed throttle elements which would slow down the rate of submissions of an agent and weed out some low-quality submissions before hitting "downstream".

I'm missing the connection to the actual capacity (or will) that the human developers have to review the submissions.

jeffbee• 2 months ago

It is an incorrect use of what was already a flawed metaphor. Pressure is isotropic. Directed pressure makes no sense, like all other fluid analogies in unrelated fields of engineering.

brookst• 2 months ago

Wait so cross ventilation, where a breeze will flow through a house if windows are open on opposite sides at a much greater rate than if windows are only open on the upwind side… isn’t really a thing?

marcosdumay• 2 months ago

The act of "making pressure" means applying a force and is completely directional.

bmm6o• 2 months ago

I took the analogy to be about the location of the pressure and not the direction. If you allow pressure to build on the input pipe when you can't accept more, the component that is upstream in the flow is able to observe that and respond. Maybe the difference is I envisioned a series of pipes and not a single one.

lucasfcosta• 2 months ago

Author here. Well noted. I do think backpressure might not be the ideal analogy/term.

It comes from previous posts I’ve come across, but I haven’t considered exactly what you mentioned. That’s on me.

ricksunny• 2 months ago

lol your comment sounds like a Claude apology

ptx• 2 months ago

Maybe it's more like shift-left testing [1]? You're trying to move some checks to earlier in the process, if I understood you correctly, and get cheaper feedback loops.

[1] https://en.wikipedia.org/wiki/Shift-left_testing

kqr• 2 months ago

It is incorrect, and that annoys me probably more than it should. Lean people have talked about this for a long time:

- single-piece flow means not making large batches of things and then sending them all downstream at once, but instead working on one thing at a time so downstream has a chance to reject before too much of the wrong thing is produced.

- autonomation (or jidoka) means giving the machine the ability to detect when something is wrong and not continue at that point.

- poka-yoke is a process that forces results to be conformant by construction.

Any and all of these terms would be better than backpressure in this context.

(This made me realise that lean people have been spending decades dealing with the problems we encounter with the new robots that write code. Half of the lean philosophy is about setting up processes and structures that have positive optionality on people's creativity, without undue requirement on their level of responsibility. That's exactly what we want for robots that write code too. We want to capture the benefits of what they do well, without suffering from their innumerable mistakes. But we can't just chastise them for making mistakes, so we have to think the way lean people do.)

SkiFreeWin3• 2 months ago

Looks like plenty of recent prior art on this:

https://pura.xyz

https://github.com/puraxyz/puraxyz/blob/main/docs/paper/main...

pshirshov5• 2 months ago

A very long post about a simple and very obvious idea with many different implementations.

The three main problems are 1) API usage is deadly expensive 2) Claude is about to make all automation very expensive 3) all the flows where a model has the initiative are strictly biased towards unwarranted stops (checkpointing).

Also, I won't call that "backpressure", there is no producer-consumer disbalance or something similar. From what I can see, the author just proposes a structured feedback loop. That's a discussion about organizational principles for system which consist of multiple unreliable but very complex components and this "backpressure" is just one of the aspects. Personally I find the viable system model framework productive as both a mental model and literal implementation guideline.

Lesser problem is that agent SDKs are bad and building a custom harness is hard.

root-parent2• 2 months ago

https://en.wikipedia.org/wiki/Back_pressure

pshirshov5• 2 months ago

The problem is not "backpressure", that's just one of the tools and there are different approaches with the same effect.

You can't express orchestration in terms of "backpressure" only, I think.

Implement-Review-Repeat loop does not involve backpressure in the strict meaning of the term.

entrope• 2 months ago

all the flows where a model has the initiative are strictly biased towards unwarranted stops

Can you elaborate on what you think causes such a bias? My experience is that Qwen3.6, Claude Sonnet 4.6 and Opus 4.6/4.7 will work as far as they can given direction and a way to test their work. My so-far limited experience with Opus 4.8 is that it does stop somewhat earlier for feedback, but in places where I am glad it is checking assumptions or where I agree with it identifying a change in scope (for example, where the following work deserves a separate commit or merge request). I would call those justified stops rather than unwarranted.

pshirshov5• 2 months ago

Ask Claude! It will quote its constitution aka soulfile. It says the constitution instructs it to perform regular checkpointing no matter what.

monkpit• 2 months ago

2) Claude is about to make all automation very expensive

Wait, what happened here??

pshirshov5• 2 months ago

They will charge use of -p and agent SDK at API rates since 14 June. So, x20..x50 price increase.

dbmikus2• 2 months ago

To stop agents from pausing for checkpointing, you can have a deterministic outer loop that re-runs until a stop condition is met.

I think teams need to be able to write nested workflows that transition between code-led and agent-led, with either supporting human-in-the-loop checkpoints.

Been iterating on what this should look like at our startup (https://www.amika.dev/). Model labs are also improving capabilities here, such as Codex's `/goal` and Claude Code's dynamic workflows[1]

The points about API usage cost still stand, but model intelligence is getting cheaper every month! No need to use the frontier model for every part of the work.

[1] https://code.claude.com/docs/en/workflows

pshirshov5• 2 months ago

It's hard to get that outer loop done, especially considering that Claude doesn't let you automate the harness anymore (it gets prohibitively expensive). Same for gemini. The only option is Codex.

/goal is a dynamic workflow itself, from what I know. Dynamic workflows do not hold the initiative (and can't use any libraries or I/O).

Dynamic workflows do not prevent checkpointing.

I don't see the actual point of your startup, it's a cheap idea - such as most LLM startups out there.

I don't see how models are getting cheaper - I clearly see the opposite trend.

2001zhaozhao• 2 months ago

You can still do backpressure and auto review workflows all you want on your CC subscription. You just have to start the task using the Claude interactive CLI, and implement your multi-agent backpressure mechanisms using Claude Code's native Subagent system with skills that tell the model to trigger them (for any subagents you want to be Claude models).

yearesadpeople• 2 months ago

If the systems invariants are well defined, and a suite of conformance + requirements tests (ensuring invariance is respected) are defined, wouldn't this be a broad - _'base case'_ - approach in general?

jon-wood• 2 months ago

This what hooks[1] are for, except hooks allow specifying criteria in certain conditions (like the agent believing it’s done and ready to hand back to the user) in a manner that the agent won’t just forget about once it’s a few turns deep, and doesn’t require triggering a whole other LLM instance to read some plain text instructions while you hope it interprets them correctly.

It absolutely makes sense to have a system in place that allows the code generated by an LLM to be automatically validated but there’s no need to resort to a non-deterministic system for these sort of deterministic pass/fail conditions.

[1] https://code.claude.com/docs/en/hooks

jarrettcoggin• 2 months ago

I was thinking the exact same thing. There are multiple places to implement hooks (git hooks, Claude hooks, etc.).

One thing I've been wondering about is how to reliably protect specific portions of the system from unexpected/unnecessary change (for example, a failing test that Claude decides to comment out or rewrite to get it to pass). My only thought for this was to automatically revert test changes during specific portions of the implementation, but that feels overly rigid and potentially prevents things like refactoring code.

cadamsdotcom• 2 months ago

Everyone looking into this and other verification should be moving away from long prompts and complex skills, and looking into hooks.

If you put all these checks in your stop hook and your git commit hook, your repo docs can tell your agent that checks will run automatically when it stops work, and it should fix any problems found.

It’s wonderful to reintroduce determinism at the QA end of your process. I find it very calming to know the agent can’t skip or forget to check its work because with hooks the checks are run by the harness.

manmal• 2 months ago

I think pi-subagents (which can form arbitrarily long chains of subagents, with up to 8 in parallel) and Claude Code‘s new workflows feature, are quite convenient abstractions that can be setup quickly.

jasonlotito2• 2 months ago

100% agree. None of this who watches the watchmen thing. Force it as a pre-commit hook. The best part there is it means you don't have to hope other people have setup there agents in the same way. It just works, every time.

xlii• 2 months ago

If that's third then I have fourth. Self plug obviously, but figured that I'd like something between smart autocomplete and an agent - an autocomplete that has wider context.

Called it rik, and it's on GitHub if anyone's interested checking it.

https://github.com/exlee/rik

EMM_386• 2 months ago

I always use a standard workflow and it has never been a problem.

- Define the task and the goal, write a short spec document (markdown is fine)

- Point the agent at it in plan mode and have it write the plan to disk with phases. Iterate on its plan if necessary here and now.

- Have each agent tackle a phase and have it update it as a living document (switch models if some phases are more difficult than others)

- Clear and repeat until done

I've never had to overcomplicate this and it's worked both on enterprise-scale projects and personal projects. I am not sure what I'm missing - if anything.

visarga• 2 months ago

I think what you are doing is good, I also have a similar workflow, but the idea here is to automate some of your manual approval work with coded tests. Since they are easy to generate, have as many as possible, think hard about what to test for, and the agent will deviate less and be more autonomous.

apf6• 2 months ago

Slowing down development is the wrong goal. I see a desire for slowness come up a lot with developers. If you pursue that goal all the way to its logical conclusion then eventually you would stop all coding completely. Which would prevent new bugs but obviously we can't do that and keep our jobs.

By all means add tons of quality gates to your SDLC pipeline. But thinking about slowness purely for the sake of slowness will not solve your problems.

apsurd• 2 months ago

If AI makes code a commodity, then why is the prescription for everyone to ship even more code even more urgently?

My gut reaction, as a professional developer, to my (previous) company's AI mandates was an instinctive "wait but..." -- it didn't logic out to me. Now that I have much more AI experience under my belt, I understand the tension, it's a superpower and net-net ok so more features and more "stuff" will be built. But it's a very hard thing to balance. It's always been a bad idea for a company to position themselves as the one with more "stuff" in it.

einpoklum• 2 months ago

In other words: Spending more tokens is all you need.

The main kind of pressure I'm feeling is the pressure of the giant AI, GPU & datacenter companies with their insane capital expenditure and circular deals, trying to get enough people to develop an expensive reliance on their service. And the more expensive, the better, so don't just pay for the LLM to code for you, have another LLM interact with the first LLM and pay double, treble, 5x or whatever. Then you can get the most refined slop.

bilbo-b-baggins• 2 months ago

Bro just rediscovered software best practices and thinks its a novel AI thing.

Fuck, we’re so cooked.

socketcluster• 2 months ago

I've been advocating for this approach for years. It's useful for any kind of data processing. You can't avoid race conditions without using some kind of queueing mechanism and you need backpressure to measure queue capacity. I built this into every aspect of https://socketcluster.io/ - From pub/sub channels, RPCs to event listeners.

jwpapi• 2 months ago

Such a fantasy, it leads to two problems.

Increased complexity of your systems. Increased pipelines of your system.

You might reduce the likelihood of errors, but at an overproportinal cost of time it takes to complete (which some might argue is irrelevant, but has the cost of human context), and with an way higher time and focus needed for all bugs that the system doesnt work.

You’ll have to fix adapt and maintain all your verification layers, because just because you set them up they are not perfect.

Your testing pipeline becomes incredible slow and you need to maintain it as well.

It’s tremendously weaker than a hands-on approach.

I’ve written this exact same article in January and since then completely switched my position.

Good luck on everyone trying this. You shuffling your own grave and waste time.

hsaliak• 2 months ago

I have a custom agent that generates patches like you will with kernel development and I review and merge those in. https://github.com/hsaliak/std_slop/blob/main/docs/mail_mode...

My agent forces this workflow by disabling modifications outside the coding step.

I added looping to this not too long ago. https://github.com/hsaliak/std_slop/blob/main/docs/mail-loop...

This gives me the best of both worlds, hand curated reviews and automation. I often get the best quality if I do both, with an agent doing a pass first.

jasonlotito2• 2 months ago

I feel like a lot of people just forget you can put this stuff in pre-commit hooks. This forces the AI to deal with issues. You don't have to hope and pray it remembers your "Pretty please, check your work" markdown file.

A pre-commit hook has been wonderful. Sure, you can add instructions, but pre-commit hooks are where you want to put the guards.

try-working• 2 months ago

I built a recursive workflow that creates its own source of truth for verifying its work: https://github.com/try-works/recursive-mode

slow_typist• 2 months ago

Who is going to write tests? But I like the fact that this approach implicitly approves of the stochastic parrot model. I mean, given enough computing power and sufficiently well made tests, I could just generate random strings of increasing length until one compiles into a program that passes all tests, mission accomplished. Like one million apes typing on one million typewriters.

lofaszvanitt• 2 months ago

Lipstick on a pig.

tim-projects• 2 months ago

I'm building a tool that automates most of this. What the author didn't even touch on is just how much AI cheats.

The more guardrails you provide the more it cheats.

AI is like a wild animal that needs to do something, and it takes a fair bit of work to corner it. And only when it's cornered and at the point of giving up, can you then offer it a way out.

If you don't do what I said, I can guarantee it's fooling you somehow.

ahstilde• 2 months ago

my entire thesis behind oro, my personal coding harness: https://github.com/mraakashshah/oro

mcint• 2 months ago

The overriding of click behavior is quite annoying. 30 years of browser user-agent behavior.

Next, Vercel, already handle this correctly. It takes special effort to violate "least surprise" here. Cmd-click on a link, should open it in a new tab.

It does appear to be an issue with SimpleAnalytics, now Adobe's,

    onclick="saAutomatedLink(this, 'outbound'); return false;"

Free debugging of how the site tweaks, breaks, the 30 year consensus web standard behavior.

Good sites, good blogs, *don't override onclick for links.* Or handle it correctly. I'll leave an issue on the github.

Between your footer, and dotfiles repo, OP does seem to appreciate standards & norms, in principle.

Backpressure is all you need

Related Stories