Detail, a San Francisco start-up that scans codebases for bugs, has launched Vuln Scans, which it says run on “the same abliterated models attackers are using to create havoc across the internet right now”. Dan Robinson of Detail, previously chief technology officer at Heap according to the company’s about page, posted on X at 23:31 UTC on 5 November that the scans have found “13k vulns so far, and the majority are already fixed”.
Robinson says Detail is delivering the scans at cost. By our arithmetic, Detail’s published pricing for its standard bug scans works out to about $2.40 per bug found and fixed in a typical codebase. Sarah Guo, founder of the venture firm Conviction, shared the launch: “Impossible to understate how important it is to find and fix all your vulns!”
Business Pill · Least privilege
An evergreen Business Engineer explainer on least privilege: giving each user or system only the access it needs. It teaches the general idea only and does not discuss Detail, OpenAI or this launch.
The key insight: As we read it, Detail is selling the opposite of OpenAI’s approach: OpenAI keeps GPT-6 Cyber for approved red teams, while Detail runs open models with their refusals removed against a sandboxed copy of any customer’s code, and says it has found 13,000 vulnerabilities so far. Its standard bug scans work out to about $2.40 per bug found and fixed, by our arithmetic.
An abliterated model is an open-weight model whose built-in refusal behaviour has been removed; a Hugging Face explainer describes the technique as one that “effectively removes the model’s built-in refusal mechanism, allowing it to respond to all types of prompts”.
Robinson’s post says the scan spins up a customer’s development environment in a sandbox, runs a large volume of abliterated-model tokens against the code, and cross-checks the software supply chain with cve/.now from Sazabi. “This is all happening within a cloud dev environment. We will not hack your prod system.” Findings can come back as pull requests that fix them, or go to an agent or ticket tracker of the customer’s choice, the post says.
Detail’s pricing page lists a free first scan, pay-per-scan credits at $0.30 each, and a team plan at $30 per contributor a month with 120 credits. It says that in a typical codebase Detail “finds and fixes a bug every 8 credits”. Its about page names angel investors including Guillermo Rauch, Olivier Pomel and Aaron Levie.
We built the vuln scanner we wanted to use but couldn’t find.
Introducing Vuln Scans in Detail!
It uses the same abliterated models attackers are using to create havoc across the internet right now.
Under the hood we are: – Spinning up your dev environment in a sandbox – Throwing an enormous amount of abliterated model tokens at trying to hack your code – Cross-checking your supply chain with cve/.now from Sazabi, to find any supply chain vulns that are exploitable in your app
This is all happening within a cloud dev environment. We will not hack your prod system.
The abliterated models are scary good. We’re finding nasty issues in every repo we scan. We’ve found 13k vulns so far, and the majority are already fixed.
We see the current moment as an all-hands-on-deck situation. Agent swarms are stringing together minor-seeming vulns across different apps to hack systems in new and unexpected ways. We all need to find and patch an enormous number of vulns, fast, or the internet is in trouble.
You don’t even have to be a target to get hacked right now. Your entire product might be a stepping stone in an attack aimed somewhere else.
We can remediate findings for you (i.e. send you PRs that fix the vulns), or you can consume them via your agent of choice, or in your ticket tracker of choice.
If you need help or advice on getting your system secure fast, please get in touch. We have helped teams of all sizes adapt to the current moment. DM me.
We’re delivering these scans at-cost, because it’s all-hands-on-deck. We hope you’ll get one today.
Dan Robinson of Detail posted the Vuln Scans launch at 23:31 UTC on 5 November. He says the scans run abliterated models against a sandboxed copy of a customer’s code, that they have found 13,000 vulnerabilities so far with the majority already fixed, and that Detail is delivering them at cost. This piece’s product details come from this post.
It’s all hands on deck to keep your IP and keep the internet running right now. Impossible to understate how important it is to find and fix all your vulns! Or at least all of them that *any model* could find. ๐
Sarah Guo, founder of Conviction, quote-posted Robinson’s launch at 00:49 UTC on 6 November: โImpossible to understate how important it is to find and fix all your vulns!โ
Two Ways to Put Attack-Grade AI on Defence
The launch came on the same day that OpenAIintroduced GPT-6 Cyber, its own vulnerability-finding model. OpenAI says the model is “initially available through the API and Codex to a select group of approved Daybreak Red customers”; its post does not give a price. OpenAI says GPT-6 Cyber helped find 81 previously unknown Firefox vulnerabilities and was behind 948 Linux exploits it published earlier.
The two approaches differ in who gets the model. OpenAI restricts its own model to approved red teams and reports how often it strays outside authorized targets: 0.07% in its honeypot tests. Detail uses open models with refusals removed and offers the scans to any customer, inside a sandbox. Robinson’s 13,000 count and OpenAI’s figures measure different things and are not comparable.
As we read the post, Detail’s pitch rests on using the same kind of models attackers use, inside a sandbox, rather than a model whose maker restricts who can use it.
As we read the pricing page, the cost per finding depends on how intensively a customer scans; the $2.40 figure is Detail’s typical-codebase rate for bug scans, not a quoted vuln-scan price.
As we read the two launches together, vulnerability finding is being sold both as a restricted premium model and as an at-cost scan on open models.
Dan Robinson of Detail, on X
“The abliterated models are scary good. We’re finding nasty issues in every repo we scan.”
Three Implications
FOR ENGINEERING TEAMS Detail says findings can arrive as pull requests that fix them, or through an agent or ticket tracker of the team’s choice.
FOR COST Detail lists $0.30 per credit and says a typical codebase yields a found-and-fixed bug every 8 credits; Robinson says vuln scans are delivered at cost.
FOR READING THE FIGURES The 13,000 count is Detail’s own and is not comparable with OpenAI’s GPT-6 Cyber figures, which measure different things.
The framework states the pattern this way: “Close the SCARCE layer (keep proprietary) + Open the ABUNDANT layer (commoditize)”.
As we read it, OpenAI keeps its cyber model closed and gated, while Detail builds its scans on open models and sells the service. The framework is a general argument about technology stacks; it does not discuss Detail, OpenAI or vulnerability scanning.
What Is Not Established
The 13,000 figure is Detail’s own and has not been independently checked; the post does not say how many repositories or customers it covers, or how severe the findings are. The pricing page does not list a separate price for vuln scans, so the $2.40 figure applies to Detail’s standard bug scans. Detail’s about page names angel investors but no funds, and we could not confirm whether Conviction has invested in Detail.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Detail has launched Vuln Scans that run abliterated open models against a sandboxed copy of a customer’s code, and says they have found 13,000 vulnerabilities so far, most of them already fixed. It is selling them at cost; its standard bug scans work out to about $2.40 per bug found and fixed, by our arithmetic. OpenAI, by contrast, restricts its GPT-6 Cyber model to approved red teams.
A note on sourcing. The product description and the 13,000 figure are from Dan Robinson’s post on X of 5 November 2026; prices are from Detail’s pricing page and the investor names from its about page, both read on 6 November. The GPT-6 Cyber figures are OpenAI’s, as reported in our 5 November piece. The $2.40-per-bug figure is our arithmetic. This piece covers the business of vulnerability scanning and gives no attack detail.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
