<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: 0x63_Problems</title><link>https://news.ycombinator.com/user?id=0x63_Problems</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Thu, 19 Nov 2026 20:14:59 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=0x63_Problems" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Gauge – Open-Source CLI to recursively improve your docs"]]></title><description><![CDATA[
<p>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.<p>The test suite eval style is useful and mostly a complement. You do want regression tests on critical features.</p>
]]></description><pubDate>Thu, 05 Nov 2026 16:18:26 +0000</pubDate><link>https://news.ycombinator.com/item?id=50380993</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=50380993</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50380993</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Gauge – Open-Source CLI to recursively improve your docs"]]></title><description><![CDATA[
<p>To clarify, this does not require the hillclimbing to be unattended, especially if you are setting up the system for the first time!<p>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.<p>It's like most AI systems in that you can have it earn more autonomy over time, as you add guidance and guardrails.</p>
]]></description><pubDate>Thu, 05 Nov 2026 16:12:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=50380931</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=50380931</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50380931</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Gauge – Open-Source CLI to recursively improve your docs"]]></title><description><![CDATA[
<p>Fair enough, maybe when we use it internally it would qualify?</p>
]]></description><pubDate>Thu, 05 Nov 2026 00:10:48 +0000</pubDate><link>https://news.ycombinator.com/item?id=50373244</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=50373244</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50373244</guid></item><item><title><![CDATA[Show HN: Gauge – Open-Source CLI to recursively improve your docs]]></title><description><![CDATA[
<p>Hi HN, I’m Evan from Gauge. We’re open-sourcing our CLI to recursively improve your docs for coding agents. The CLI includes a local execution backend, so you don't need an account to try it. Explainer video here: <a href="https://youtu.be/zhLUlIKB1Mg" rel="nofollow">https://youtu.be/zhLUlIKB1Mg</a><p>The CLI supports a hillclimbing loop:<p>- set up tasks for coding agents that would need your docs, like 'set up Gauge CI checks'<p>- run the tasks<p>- analyze session transcripts for friction or failures<p>- create variants of your docs to resolve the issues<p>- re-run the tasks with the doc variants injected (transparently, at the same hostname)<p>- repeat until you have a variant that resolves the issues<p>The same loop supports hillclimbing your Skills, CLI, and other agent-facing surfaces.<p>We started out 2 years ago building open-source developer tools (like Tach <a href="https://github.com/tach-org/tach" rel="nofollow">https://github.com/tach-org/tach</a>), then pivoted to AI Search optimization, and are now focused on the intersection of the two spaces. We were already working with awesome devtool companies, and they were the first ones to ask us how to optimize for Claude Code and Codex. We realized this would be something that eventually every company needs to do. Our CLI provides primitives for docs/devrel/eng teams to measure, analyze, and improve how coding agents use their product.<p>The design boils down to two components: VMs and a network gateway.<p>The VMs are straightforward, we want the agents to be highly privileged within their environment so that they are representative of developers' coding agents. Each VM generally comes with the agent harness itself, language toolchains, and a git repository.<p>The network gateway is where we terminate TLS and swap in variants of your docs transparently. The gateway also handles credential replacement, so agents see placeholders and can't exfiltrate keys.<p>The CLI comes with a 'local' execution backend (inference is still remote). For VMs, it uses Docker containers, and the network gateway is mitmproxy running in a sidecar container for each session.<p>We've used the hillclimbing loop internally to find and fix coding agent issues like mistaken pricing, inaccurate product limitations, and repeated workarounds. We saw coding agents report that Gauge only supports Claude Code and Codex, traced it back to the web pages they fetched, and fixed the behavior (we support Pi, OpenCode, and Antigravity as well). For Clerk, we helped shift coding agents from their legacy 'keyless' mode to their new 'clerk init' command. They were able to validate a pre-release change to their SDK's error messages using Gauge, then saw a 3x increase in coding agents using 'clerk init' post-deploy.<p>Due to the 'static' nature of AI models, we've realized that closed-loop optimization techniques are now possible for essentially all software products. The same agent can be simulated over and over, meaning that a high-fidelity testing environment is all you need. Achieving the best results boils down to high-quality analysis (finding generalizable friction and objections) and paired experiments (vary only the content, measure the downstream change in behavior).<p>The repo is here: <a href="https://github.com/gauge-sh/gauge-cli" rel="nofollow">https://github.com/gauge-sh/gauge-cli</a> and you can start using the CLI with: npm i -g @withgauge/cli<p>We also have documentation and a quickstart here: <a href="https://docs.withgauge.com/agents/guides/cli">https://docs.withgauge.com/agents/guides/cli</a><p>If you point a coding agent at those pages, it should guide you through the initial setup - let us know if it doesn't! The intention is for your agent to run some sensible evaluations and find at least one point of friction you can fix.<p>We'd love to know from HN: Has anyone tried to DIY this locally, or built something in-house at their company? Is this an improvement? If not, what would need to change?</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=50371090">https://news.ycombinator.com/item?id=50371090</a></p>
<p>Points: 43</p>
<p># Comments: 10</p>
]]></description><pubDate>Wed, 04 Nov 2026 20:31:30 +0000</pubDate><link>https://github.com/gauge-sh/gauge-cli</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=50371090</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50371090</guid></item><item><title><![CDATA[Show HN: Gauge CLI – Open-Source Hillclimbing for Agent Experience]]></title><description><![CDATA[
<p>Hi HN! We're releasing an open-source CLI for hillclimbing Agent Experience. That means optimizing your agent-facing surfaces (Skills, CLIs, SDKs, and docs) so that your users' coding agents can easily use your product. Pointing your own agent at the repo is enough to get started. <a href="https://github.com/gauge-sh/gauge-cli" rel="nofollow">https://github.com/gauge-sh/gauge-cli</a><p>We've found that power users generally want to hand a CLI/MCP to their agents and get started from the terminal. They rarely want to log in and use the web UI themselves. We think developer tools must be agent-friendly to remain relevant.<p>Here are some common findings that Gauge identifies:<p>- agents get confused with CLI/SDK behavior and start source diving
- interactive commands block agents
- agents use deprecated/old CLI commands
- agents don't use Skills when you'd expect
- agents confidently assert limitations that don't exist
- inconsistent terminology confuses agents
- agents can't figure out how commands fit together into a workflow<p>Locally, the Gauge CLI uses Docker containers to run agent harnesses (you can bring your own subscriptions). It finds where agents get confused or stuck and tests changes using your pre-release artifacts (Skills/CLI etc). We normalize events from Claude Code, Codex, Pi and OpenCode into standardized transcripts, and store them locally, alongside structured information in SQLite so that your agent can easily analyze the data. This enables automatic hillclimbing toward a great Agent Experience.<p>We also offer hosted compute that enables more features, built on our custom Gateway (for credential brokering and transparent, per-session response rewriting). Hosted compute enables many more concurrent sessions, easier collaboration with teammates, a GitHub App for PR reviews, sessions with logged-in MCPs/APIs, and most importantly it lets you close your laptop while the sessions actually run.<p>You don't need a Gauge account to run the full loop end-to-end though. We expect that local execution with the CLI is enough to find and fix several agent experience issues in a typical project.<p>You'll be able to validate fixes for these issues locally, before publishing new versions (we inject pre-release Skills and code artifacts into the sessions).<p>We would love your feedback!</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=50170201">https://news.ycombinator.com/item?id=50170201</a></p>
<p>Points: 5</p>
<p># Comments: 0</p>
]]></description><pubDate>Tue, 20 Oct 2026 18:01:22 +0000</pubDate><link>https://github.com/gauge-sh/gauge-cli</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=50170201</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=50170201</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Ax-check.com – Can agents use your product?"]]></title><description><![CDATA[
<p>You can reproduce that finding directly from python:<p>```<p>>>> request.urlopen("<a href="https://anc.dev" rel="nofollow">https://anc.dev</a>")<p>[...traceback info...]<p>urllib.error.HTTPError: HTTP Error 403: Forbidden<p>```<p>In this case (<a href="https://agents.withgauge.com/p/runs/40be763c-4678-4677-88df-b5ff07163b62">https://agents.withgauge.com/p/runs/40be763c-4678-4677-88df-...</a>) 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.</p>
]]></description><pubDate>Fri, 18 Sep 2026 22:39:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=49761212</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=49761212</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49761212</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Ax-check.com – Can agents use your product?"]]></title><description><![CDATA[
<p>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</p>
]]></description><pubDate>Fri, 18 Sep 2026 20:18:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=49759685</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=49759685</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49759685</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Ax-check.com – Can agents use your product?"]]></title><description><![CDATA[
<p>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.</p>
]]></description><pubDate>Fri, 18 Sep 2026 19:00:35 +0000</pubDate><link>https://news.ycombinator.com/item?id=49758696</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=49758696</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49758696</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Ax-check.com – Can agents use your product?"]]></title><description><![CDATA[
<p>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.<p>For a lot of products it can just be an extension of the free tier I think too.</p>
]]></description><pubDate>Fri, 18 Sep 2026 18:58:12 +0000</pubDate><link>https://news.ycombinator.com/item?id=49758658</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=49758658</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49758658</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Ax-check.com – Can agents use your product?"]]></title><description><![CDATA[
<p>Are you mainly concerned about scraping? or anti-agents in general?</p>
]]></description><pubDate>Fri, 18 Sep 2026 18:52:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=49758589</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=49758589</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49758589</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Ax-check.com – Can agents use your product?"]]></title><description><![CDATA[
<p>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!</p>
]]></description><pubDate>Fri, 18 Sep 2026 18:49:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=49758558</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=49758558</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49758558</guid></item><item><title><![CDATA[Show HN: Ax-check.com – Can agents use your product?]]></title><description><![CDATA[
<p>I'm the co-founder of Gauge, and I built ax-check.com to quickly test how well coding agents can onboard to your product.<p>You'll get a scorecard, specific suggested fixes, and three full coding sessions that show how agents read your site and use your product.<p>I built this because similar checks were too noisy. Most suggested obscure technical changes that don't actually make a difference in agent experience (or AX, hence ax-check.com).<p>This check starts by using DeepSeek 4.1 Flash to try to find key information about your product, starting from the homepage. In actual agent traffic data, we've seen that the key pages are the homepage, llms.txt, pricing, and the docs site (by traffic volume, and by influence), so we focus on those and ignore the rest. We also find that content negotiation for Markdown is legitimately helpful for agents to complete tasks faster and find what they're looking for, so the scan tests that your key pages can serve Markdown.<p>The other key piece is that we run actual coding agents in sandboxes, and have them try to onboard to your product. You can see the full trace and watch it happen live (we kick it off fresh when you enter a new site). We surface interesting findings like hallucinated URLs, inaccurate docs instructions, or product confusion.<p>It also detects whether the agents could complete a fully working onboarding autonomously, without being blocked by a login wall. This is still controversial, but I think finding ways to let agents safely onboard autonomously is going to be table stakes within a year for developer tools in particular.<p>The whole site is agent-friendly itself! You can generally just talk to your coding agent about ax-check.com and it can do the rest. Would really appreciate any feedback to make this useful.</p>
<hr>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=49744416">https://news.ycombinator.com/item?id=49744416</a></p>
<p>Points: 38</p>
<p># Comments: 41</p>
]]></description><pubDate>Thu, 17 Sep 2026 18:08:02 +0000</pubDate><link>https://www.ax-check.com/</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=49744416</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49744416</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Research acceleration: The view inside OpenAI"]]></title><description><![CDATA[
<p>I think it's only recursive from the perspective of the humans, i.e. they design Astra, which itself as part of its deployment designs Galactica, etc.<p>So humans develop things one after the other, but when the thing itself starts developing new things, those are happening 'recursively' in its scope.</p>
]]></description><pubDate>Sun, 06 Sep 2026 18:31:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=49589418</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=49589418</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=49589418</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Ask HN: Who is hiring? (January 2026)"]]></title><description><![CDATA[
<p>Gauge | Founding Engineer | San Francisco, CA | ONSITE | Full-time<p>Gauge helps companies show up in more AI answers. Search traffic is shifting away from Google into AI chat, and companies are figuring out how to adapt right now. Our system delivers real-time analytics derived from millions of LLM answers, while providing an intuitive, powerful UX. We also generate optimized articles to improve our customers’ visibility, and offer an agentic chat interface over their data (think ‘Why was there a spike on 12/10?’ or ‘Which of my pages should I refresh?’).<p>I'm Evan, the CTO, and we're looking for engineers with strong programming fundamentals, creative problem solving skills, and a high-agency working style. You might be a good fit if you are an ex-founder, ex-founding engineer, or have worked at a Seed/Series A startup.<p>We primarily use Django, React, Postgres, Celery, and Clickhouse.<p>Contact me with your resume (and mention HN) if you're interested: evan[at]withgauge.com<p>More details/JD also here: <a href="https://www.ycombinator.com/companies/gauge/jobs/qemTjv7-founding-engineer">https://www.ycombinator.com/companies/gauge/jobs/qemTjv7-fou...</a></p>
]]></description><pubDate>Fri, 02 Jan 2026 19:48:51 +0000</pubDate><link>https://news.ycombinator.com/item?id=46468606</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=46468606</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46468606</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Comparing product rankings by OpenAI, Anthropic, and Perplexity"]]></title><description><![CDATA[
<p>Very cool!<p>Yup I definitely see confusion in our responses around the product and brand names. We do another pass through an LLM specifically aimed at ‘canonicalizing’ the names, but we’ll need to get more sophisticated to catch most issues.<p>In that case you mentioned, the brand confusion is what accounts for the top three omission for QBO. Both OpenAI and Perplexity rank it #1, but Anthropic ranks the slightly different “Quickbooks” product as #1. Our overall ranking prioritizes products that appear in all three responses, so both are dropped down.</p>
]]></description><pubDate>Wed, 09 Apr 2025 19:34:22 +0000</pubDate><link>https://news.ycombinator.com/item?id=43636716</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=43636716</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43636716</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Comparing product rankings by OpenAI, Anthropic, and Perplexity"]]></title><description><![CDATA[
<p>No specific reason, just started with these three, will add Gemini soon!</p>
]]></description><pubDate>Wed, 09 Apr 2025 16:37:11 +0000</pubDate><link>https://news.ycombinator.com/item?id=43634085</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=43634085</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43634085</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Comparing product rankings by OpenAI, Anthropic, and Perplexity"]]></title><description><![CDATA[
<p>It's from previous searches actually, we have an 'enrichment' step after the initial rankings come back which helps with semantic deduplication and tries to give us a canonical website domain. We store the Product and tag all matching rankings: <a href="https://productrank.ai/product/microsoft" rel="nofollow">https://productrank.ai/product/microsoft</a> and use a 3rd party to map website <-> brand logo.</p>
]]></description><pubDate>Wed, 09 Apr 2025 16:36:38 +0000</pubDate><link>https://news.ycombinator.com/item?id=43634079</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=43634079</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43634079</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Comparing product rankings by OpenAI, Anthropic, and Perplexity"]]></title><description><![CDATA[
<p>We are using sonar-pro</p>
]]></description><pubDate>Wed, 09 Apr 2025 16:34:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=43634041</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=43634041</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43634041</guid></item><item><title><![CDATA[New comment by 0x63_Problems in "Show HN: Comparing product rankings by OpenAI, Anthropic, and Perplexity"]]></title><description><![CDATA[
<p>These are structured results from explicitly asking the LLM for a ranking in the given category, and we provide guidance in the prompt telling the LLM to 'use best judgment' when the topic doesn't clearly include products.<p>Also we include the 'key features' from each answer - you can see this by clicking the cell containing the rank (e.g. '1st' in the Anthropic column)<p>In this case, Anthropic said of 'Death during sleep':<p>Anthropic Analysis for Death During Sleep<p><pre><code>    Painless and unaware experience
    No anticipatory anxiety
    Common with certain cardiac conditions
    Often described as 'peaceful'
    No suffering</code></pre></p>
]]></description><pubDate>Wed, 09 Apr 2025 15:48:15 +0000</pubDate><link>https://news.ycombinator.com/item?id=43633449</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=43633449</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43633449</guid></item><item><title><![CDATA[How to Build Hot Module Replacement in Python]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.gauge.sh/blog/how-to-build-hot-module-replacement-in-python">https://www.gauge.sh/blog/how-to-build-hot-module-replacement-in-python</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=43494949">https://news.ycombinator.com/item?id=43494949</a></p>
<p>Points: 3</p>
<p># Comments: 0</p>
]]></description><pubDate>Thu, 27 Mar 2025 15:54:49 +0000</pubDate><link>https://www.gauge.sh/blog/how-to-build-hot-module-replacement-in-python</link><dc:creator>0x63_Problems</dc:creator><comments>https://news.ycombinator.com/item?id=43494949</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43494949</guid></item></channel></rss>