Active work should mostly be hillclimbing, or at least more exploratory. The search space is very large and you typically find a lot more low-hanging friction than you'd think by just running a broad battery of sessions. And I'd pair active hillclimbing work with feature launches/product work - essentially part of releasing new things is QA-ing them against coding agent workflows.
The test suite eval style is useful and mostly a complement. You do want regression tests on critical features.
The search space being large feels like a fundamentally hard and costly problem to solve for. Where do you recommend folks start given such a high number of possible scenarios?
To clarify, this does not require the hillclimbing to be unattended, especially if you are setting up the system for the first time!
This mainly automates the process that you are doing by hand. If you have your own definition of friction and priorities, you can tell this to your agent and it will fit into the CLI loop. You can review any change that is suggested, and spot-check transcripts if you feel the need.
It's like most AI systems in that you can have it earn more autonomy over time, as you add guidance and guardrails.
In this case (https://agents.withgauge.com/p/runs/40be763c-4678-4677-88df-...) the agent wrote a python script to test the site, and hit the error above. We don't have a full filesystem diff viewer in ax-check.com, although we do have that in the Gauge agents product itself.
Interesting, thank you. I'll check whether a setting on CF might be interfering. The site checks headers for routing and response types, but it shouldn't affect access.
which URL is it? it's kind of a deep assumption since we're evaluating on ease-of-navigation from the homepage, and looking for /llms.txt etc. but there should be a way for me to add support
Each product is different (how much does it cost to serve a marginal user, how risky is it). But generally I think allowing ephemeral accounts like Cloudflare, or serving simulated/static accounts before a human claim can both work.
For a lot of products it can just be an extension of the free tier I think too.
Agents are the new search engine - a majority of people are starting to get recommendations and answers from them. So asking this is like asking "what is the benefit of making any resource more available to search engines?"
The answer is discoverability. which is pretty critical for any business
I've heard CTOs saying they don't consider buying any product that is not AI agent friendly. Which means they must come with AI skills, MCP Servers and so on that agents can use it to do stuff. You may not like it, but that's almost certainly going to be the future.
agents are becoming the de facto standard interface for everything, whether you like it or not, so it depends on the intent you have for a project, or "product" as OP phrased it, which has a more contextual implication that the intent is to get others to buy it
ironically, I did sell a kirby many many years ago
the idea is likely way more about business purchases and SaaS than consumer products, though I did have an agent do deep research to give me options when deciding on a new mattress
Behind an agent is a human that might use the product. I think customers/consumers will demand that their personal/subscription agent can access the product.
As someone that works for a company that gets a 100% score on ax-check, all the effort I've put into making it accessible for agents has not 10xed the growth numbers like I was told it would.
The Ai growth looks more like the very-long-term 2% (technically productivity growth from innovation), avoid assuming the hypesters are even close to accurate in their predictions
The ROI is for we the users, not the companies, only issues is that the agent's attention is way easier to hijack, so you have to maintain a different kind of attention when using them and never blindly trust their outputs.
The initial step of the scan looks at homepage content to try to figure out where to navigate next, it's getting confused by all the different products linked. Will fix this!
The test suite eval style is useful and mostly a complement. You do want regression tests on critical features.
reply