Oblique

Oblique

@oblique.security

Scale access securely and automatically

34 followers5 following37 posts

Pinned

Identity management has quietly become the primary security perimeter. But it's a mess — identity requires constant manual work that security teams burn out from. At Oblique, we're helping organizations make their access controls actually maintainable. Full post: oblique.security/blog/identit...

oblique.securityIdentity management is harder than it should be | ObliqueIdentity management is surprisingly hard, as access controls change constantly, and getting them right requires context. We founded Oblique to work on impactful security problems.

If you're the first (or only) security person at your startup, you're probably juggling infrasec, appsec, corpsec, GRC, and maybe even physical security. Come meet a room full of people in the same boat on Wednesday in SF, for our Day One Security during SF Tech Week: partiful.com/e/XJr1dx0LSC...

partiful.comDay One Security #SFTechWeek | PartifulNot sure what you should be doing about security? Hear from people who've already worked it out. Join us for lightning talks from first security hires and security professionals, then stay to mingle ...

By limiting "sign in with Google" permissions, you can control what apps users access, and what OAuth grants they give those apps. You shouldn't block sign in — users will find another way — but you can block additional OAuth grants. More in our blog post: oblique.security/blog/unconfi...

oblique.securityThe wrong way to block "Sign in with Google" | ObliqueYou can block your users from signing into apps with their Google account, and granting access to their inbox. But not knowing what they're using is worse.

Our co-founder @ericchiang.bsky.social had fun with Claude finding vulns in SAML implementations. Using a hacking harness, what he found was full authentication bypasses in multiple libraries, tired maintainers, and AI-automated fixes. oblique.security/blog/hacking...

oblique.securityHacking SAML with Claude Code | ObliqueAfter many years of complaining about how insecure SAML is, I decided to try to prove it by using Claude Code to hack every SAML implementation I could find.

We recently completed our first SOC 2 Type II audit, covering the Security and Confidentiality trust services criteria. Our cofounder @mayakaczorowski.com shares what's still hard: manual access reviews, scattered evidence, and complying with bygone requirements: oblique.security/blog/soc2-re...

oblique.securityWhat's still annoying about SOC 2 in 2026 | ObliqueWe completed our first SOC 2 audit. We still suffered through manual access reviews, scattered evidence, and proving compliance with bygone requirements.

Our background jobs were inexplicably slow. We looked at existing tracing, added OTel, and hypothesized. It looked like a database problem: multi-second SQL queries, connection pool exhaustion, lock contention candidates. The fix that @jbrandhorst.com found turned out to be one line of YAML 😅

A coworker disappearing into a cave and coming back with 12,000 lines of code you have to review is cause for a heart attack. But, how do you handle the ever-increasing volume of code reviews? Our engineering team wrote down how we navigate code reviews at Oblique: oblique.security/blog/how-obl...

oblique.securityHow Oblique handles code review etiquette | ObliqueCode reviews aren’t just about ensuring engineering quality, they’re also about building team culture. Focus on communication and understanding rather than nit-picking style quirks.

Great security doesn't have to cost a lot. Our security stack costs us nothing. By using free tiers or open-source versions of popular tools (Semgrep, Truffle, RunReveal, and Sublime), we're doing more than just checking a box. @mayakaczorowski.com gives an overview: oblique.security/blog/securit...

oblique.securityThe $0 security stack | ObliqueAs part of our initial SOC2 audit, we wanted to put in place actually useful security tools. Luckily, in 2026, world-class security costs literally nothing.

Access controls should map to how your organization works — not the other way around. Organizations generally grant access based on department, based on reporting chain, and based on projects. Use what works for your org. oblique.security/blog/org-str...

oblique.securityFit access controls to your org, not the other way around | ObliqueGroups used for access controls can be based on department, reporting chain, or projects. The right answer is whatever maps best to how your org actually works.
View full thread

In other parts of security, we’ve learned that fixing classes of problems is the only way we can address them for good — but we haven’t had that mindset shift yet in access management. We need better controls, so that we can get to fewer tickets, not faster tickets. oblique.security/blog/access-...

oblique.securityAccess requests are a bandaid, not a fix | ObliqueIT teams are overwhelmed with never-ending access requests. Getting off the identity treadmill means getting to fewer tickets over time, not faster tickets.

It’s 2025 and many teams still can’t reliably enforce strong authentication across their app stack. That’s the real SSO tax: not paying to have SSO, but paying to enforce it. Read more about how we’re approaching practical enforcement at Oblique: oblique.security/blog/real-ss...

oblique.securityThe real SSO tax | ObliqueThe SSO tax shouldn't be about having SSO — it should be about enforcing it. The value of SSO is to centrally manage access and require strong authentication.

What you really want to control access to is data, not systems — so why are we stuck thinking in systems? Our cofounder @mayakaczorowski.com shares what she learned researching tiered controls for our latest report on Modern Access Controls.

Authentication failures from the last five years at Okta, Snowflake, and Twitter show very similar attacks, from credential theft, to MFA bypass, to session hijacking. Dive deeper into these incidents and avoid repeating the same mistakes: oblique.security/blog/authn-f...

oblique.securityWhat we can learn from real-world authentication failures | ObliqueRecent breaches at Okta, Snowflake, and Twitter help us learn how to prevent authentication failures like credential theft, MFA bypass, and session hijacking.

Don't rely on managers for access approvals — they don't work, for either security or speed. Instead, get approvals from app owners who actually understand the systems and risks, and automate approvals that are always granted.

The biggest scaling challenge for IT and security teams isn't technical — it's organizational. When you're managing access for thousands of employees and hundreds of applications, you need to know: who owns what? Read more in our latest post: oblique.security/blog/delegat...

oblique.securityDelegate authority to those with context | ObliqueBusiness teams have context for access decisions but lack authority. Delegate to those closest to the resources by defining clear ownership for each app.

If you're interested in learning more about what's happening in the IAM market — and who's competing with Okta and why — then you should read our cofounder @mayakaczorowski.com's latest post.

Maya Kaczorowski@mayakaczorowski.com

Wrote up some thoughts for how Okta is getting squeezed from all sides: squished by Rippling at the bottom, Msft at the top, and other markets and startups in the middle. I hope you love some red string diagrams 🤣 ventureinsecurity.net/p/the-unbund...

Your job title makes a bad RBAC role: what access does a Chief Happiness Officer need, anyways? A role in RBAC should represent what someone actually does in your environment. Your job title is your position, not your job function. Read more in our latest blog post: oblique.security/blog/rbac-ro...

oblique.securityWhy your RBAC roles aren't actually roles | ObliqueA role in RBAC should represent what someone actually does in your environment. Your job title makes a bad RBAC role: it's your position, not your function.

Comms groups map to how people actually work, and often, access groups don't (but they should). But comms groups always become access groups. It's not a matter of if, but when. Read more in our latest post: oblique.security/blog/comms-a...

oblique.securityComms groups inevitably become access groups | ObliqueComms groups map to how people actually work, and often, access groups don't (but they should). Comms groups always become access groups. It's not a matter of if, but when.

Check out the latest from our cofounder @ericchiang.bsky.social to learn about a neat Go type trick to avoid query injection in SQL builders.

Eric Chiang@ericchiang.bsky.social

Wrote about a fun @golang.org type trick where APIs can force clients to pass string constants as arguments. Happens to be _extremely_ useful for SQL builders! oblique.security/blog/injecti...

Over the past 60 years, we've gone from reusing the same password everywhere to advanced biometric authentication like FaceID. Dive into the history of authentication in just 2 minutes!

Authentication has evolved from simple passwords to federated systems with passwordless logins, with a constant push and pull to balance security and usability. Deep dive into the evolution of authentication in our latest blog post! oblique.security/blog/history...

oblique.securityThe evolution of authentication, from passwords to passkeys | ObliqueAuthentication has evolved from simple passwords to federated systems with passwordless logins, continuously balancing security and usability.

Instead of minting long-lived API keys, you can use GitHub Actions' OpenID Connect support for workload identity. Here's how we authenticate config-as-code workflows in Oblique without secret management headaches. Better security + Better developer experience 💟 oblique.security/blog/github-...

oblique.securityAuthenticating GitHub Actions without API keys | ObliqueInstead of minting long-lived APIs keys and warning users “keep this secret,” let's use GitHub Action's OpenID Connect support instead.

IT teams are afraid of removing access — what if something breaks? Even if you don't know why someone has access, you should be able to figure out if they're using it. Removing unused access isn't risky — never removing access is. Read more in our latest blog post: oblique.security/blog/chester...

oblique.securityChesterton's fence doesn't apply to access controls | ObliqueIT teams are scared to remove access they don't understand, leading to sprawling entitlements. Removing unused access isn't risky — never removing access is.

You are up to date.