Security
The security basics AI-built apps skip — secrets, logins, row-level security, bot abuse — and the isolation that keeps untrusted code contained.
140 posts · page 1 of 6
What Is SSO (Single Sign-On)? How It Works and When Your App Needs It
Single sign-on lets people log in once with one company account — Google Workspace, Microsoft Entra, Okta — and get into many apps without separate passwords. How SSO works, SAML vs OpenID Connect, why business customers demand it, and how to add it to your own app.
What Is OAuth? "Sign in With Google" and Delegated Access Explained
OAuth is the standard that lets one app access your data in another app without ever seeing your password — it's behind "Sign in with Google" and "Connect your GitHub account". How the OAuth flow works step by step, what scopes and tokens are, OAuth vs OpenID Connect, and the mistakes to avoid.
What Is Base64? Encoding Explained (and Why It Isn't Encryption)
Base64 turns any data — images, files, binary keys — into plain text made of 64 safe characters, so it can travel through systems that only handle text. How it works, where you'll see it (data URLs, JWTs, API keys, email), how to encode and decode it, and why it provides zero security.
What Is an SSL Certificate? (And Do You Need to Pay for One?)
An SSL certificate (really a TLS certificate) proves your website is who it says it is and lets browsers encrypt the connection — it's what puts the padlock and https:// in front of your site. What's inside one, how browsers check it, free vs paid certificates, expiry, and the errors you'll see when it goes wrong.
VPN vs Proxy: What's the Difference and Which Do You Need?
A VPN encrypts traffic routed through its tunnel; a proxy forwards traffic selected by an app. How full and split tunnels work, what VPNs and proxies do and don't hide, reverse proxies vs forward proxies, and which one fits privacy, work access, scraping or testing.
"User Is Not in the Sudoers File": What It Means and How to Fix It
"username is not in the sudoers file. This incident will be reported." means your Linux user isn't allowed to run commands as root with sudo. How to add a user to the sudo (or wheel) group from root, edit sudoers safely with visudo, recover when you've locked yourself out, and the safer alternatives to giving out sudo.
SBOMs and Signed Images: Software Supply Chain Security With Syft and Cosign
How to know what's in your container images and prove they came from your CI: generating SBOMs with Syft or BuildKit attestations, scanning them for vulnerabilities, signing images keylessly with Cosign and GitHub Actions OIDC, verifying by identity, adding SLSA provenance, and enforcing it at deploy time.
Nginx Rate Limiting: limit_req, Bursts and Protecting Your Login Page
How to rate-limit requests in Nginx with limit_req_zone and limit_req: choosing a key and rate, what burst and nodelay really do, returning 429 instead of 503, separate limits for login and API routes, getting the real client IP behind Cloudflare, connection limits, and testing it.
nftables vs iptables: The Linux Firewall, Then and Now
nftables is the successor to iptables in the Linux kernel: one tool for IPv4, IPv6 and more, sets and maps, atomic rule updates, and a cleaner syntax. How the two relate (including iptables-nft), a complete server ruleset, translating rules, and the Docker, UFW and firewalld interactions that trip people up.
Next.js September 2026 Security Release: Upgrade to 16.3.8 or 15.5.27
Next.js patched seven vulnerabilities on 30 September 2026 in v16.3.8 and v15.5.27, after an out-of-band critical fix on 22 September. A high-severity SSRF in image optimisation, several cache-poisoning and cache-leak issues for SSG/ISR and 'use cache', and a dev-server MCP endpoint that leaked project data to malicious websites. Who's affected and what to do.
MemTensor npm and PyPI Packages Compromised With a Self-Spreading Credential Stealer
On 23 September 2026 attackers published malicious versions of @memtensor/memos-cloud-openclaw-plugin on npm and MemoryOS on PyPI carrying a Go implant ("sckit") that steals developer and CI credentials and spreads into other packages its victims publish. Affected versions, how it works, indicators, and the cleanup steps.
Lovable Now Uses Free and Pro Plan Data for AI Training Unless You Opt Out
Since 9 September 2026, Lovable may use customer data from Free and Pro plans to train AI models unless users opt out in Account settings → Privacy. Business and Enterprise workspaces are excluded by default. What that covers, how to opt out, and what to think about if your app handles client or customer data.
Lovable Moves Into the Enterprise: Microsoft Entra Sign-In, Azure and Power BI Connectors
Through September 2026 Lovable added Microsoft sign-in for apps, Entra gallery listing, SSO management, Power BI and Azure connectors, and on 30 September announced that apps built by a team can run inside the company's Microsoft tenant with Entra sign-in and IT controls. Plus EU inference and IP-allowlisted API keys for Enterprise. What it means for internal tools.
Let's Encrypt Moves to 64-Day Certificates in February 2027
Let's Encrypt announced on 7 October 2026 that certificates issued from 10 February 2027 will last 64 days by default instead of 90, with staging switching on 14 October 2026. Authorization reuse drops from 30 to 10 days. Why it's happening, who it breaks (hard-coded renewal schedules), and how to check your setup.
keyv, cacheable-request and 400+ npm Packages Hit by a "Mini Shai-Hulud" Worm
In August 2026 attackers compromised a maintainer's GitHub account and published malicious versions of keyv, cacheable-request, cache-manager and other widely used npm packages. The Shai-Hulud-style worm stole cloud, CI and AI-tool credentials, persisted through Claude Code hooks and VS Code tasks, and spread to over 400 packages. How it worked and how to clean up.
GitHub Enforces Minimum Self-Hosted Runner Versions on Enterprise Cloud
Since 29 September 2026, GitHub Enterprise Cloud enforces minimum versions for self-hosted Actions runners: runners below 2.329.0 can't register or re-register, and runners below a higher job-execution minimum stop running jobs. Who's affected, how to check your fleet with the new runner-deprecation REST API, and how to keep runners updated.
GitHub Actions Will Block pull_request_target by Default in Public Repos From 2 November
GitHub made workflow execution protections generally available on 17 September 2026: actor and event rules that control who and what can trigger workflows. For public repositories, a default rule disabling pull_request_target starts in evaluate mode and will be enforced on 2 November 2026. Why, and what maintainers should do now.
GitHub Actions Caching: Faster CI With setup-node, actions/cache and cache-mode
How to make GitHub Actions workflows faster with caching: the built-in cache option in setup-node and setup-python, actions/cache for anything else, good cache keys, Docker layer caching, the new cache-mode setting that limits cache access to prevent cache poisoning, and why caches miss.
How to Generate a Secret Key (for JWTs, Sessions and APIs)
Apps need long random secrets for signing sessions and JWTs, encrypting data, and verifying webhooks. How to generate a strong one with openssl rand, Node.js, Python or a password manager, how long it should be, where to store it, and the mistakes that make secrets guessable.
ERR_SSL_PROTOCOL_ERROR: What It Means and How to Fix It
ERR_SSL_PROTOCOL_ERROR means the browser couldn't complete a secure HTTPS connection because the server's response wasn't valid TLS. The usual causes — HTTPS pointed at a port speaking plain HTTP, a missing certificate, localhost over https, Cloudflare SSL modes — and how to fix each.
.dockerignore Explained: Faster Builds, Smaller Images, No Leaked Secrets
A .dockerignore file keeps files out of the Docker build context: node_modules, .git, .env files, build output. Why that makes builds faster and images smaller, how it prevents secrets ending up in images, the syntax and negation rules, a ready-to-use template, and how to check what's being sent.
Cloudflare Plans a Free Public Certificate Authority With Post-Quantum Merkle Tree Certificates
On 29 September 2026 Cloudflare announced it intends to become a public certificate authority, issuing free classical TLS certificates and post-quantum Merkle Tree Certificates (MTCs) via ACME, only to clients that support ARI. Why post-quantum certificates need a new format, what's still pending (root programs, a GlobalSign root deal), and what it means for site owners.
Cloudflare Protected Quick Tunnels: Share Localhost Only With the Emails You Choose
Cloudflare's Protected Quick Tunnels (Birthday Week 2026) add an --allowed-mail flag to cloudflared 2026.9.3+, so a quick tunnel to your local dev server only admits listed email addresses or domains, verified with a one-time PIN — no Cloudflare account needed on either side. How it works, its limits, and the coding-agent angle.
Claude Code Hooks Can Now Fail Closed — and Prompt Hooks Got a Security Fix
Claude Code 2.1.295 (8 October 2026) adds onFailure: "block" for command and HTTP hooks, so a hook that crashes or times out blocks the action instead of silently allowing it. 2.1.294 fixed prompt and agent hooks written as instructions sometimes allowing what they should block. What to update.