SShortSingh.
TopTechnologyAI & MLStartupsProgrammingScienceBusinessFinanceCrypto & Web3World
Back to feed
Programming

Synapse Shield v0.10.0 Extends Behavioral Bot Detection to Mobile Browsers

0
·14 views

Synapse Shield, an open-source self-hosted behavioral biometrics engine, has released version 0.10.0 to bring its bot detection to mobile devices. The engine scores touch, mouse and keystroke dynamics using only relative millisecond timings, avoiding both CAPTCHAs and personally identifiable information. The update adds a Mobile SDK v2.0 for vanilla JS, React and Vue, along with changes to the Python engine and native Rust core. Developers addressed five false-positive scenarios that desktop-oriented rules had misflagged on mobile traffic, including single taps, IME keyboards and zero-radius iOS WebView touches. The release is a web SDK for mobile browsers, PWAs and WebViews, and does not yet include a native Swift or Kotlin SDK, with real-world device validation still pending.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

Moving procedures from CLAUDE.md into Claude Code skills for teams

Team CLAUDE.md files often grow past the point where agents follow them reliably, and Claude Code's documentation recommends keeping each file under 200 lines. Much of that content, such as release checklists and migration steps, is only needed occasionally rather than in every session. Skills address this by loading instructions only when relevant, and each is a folder containing a SKILL.md file with frontmatter and instructions. The article suggests a test for deciding placement: content needed in every session belongs in CLAUDE.md, while task-specific procedures belong in skills. Skills can be saved at project, personal, or enterprise level, and name conflicts are resolved with enterprise taking precedence over personal, and personal over project.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Quantum computing could threaten blockchain cryptography, prompting post-quantum migration plans

An educational guide examines whether quantum computers pose a risk to blockchain security. It explains that quantum machines use qubits to factor large numbers far faster than classical computers, which could undermine the encryption that current blockchains rely on. The article outlines steps to address this, including assessing vulnerable signature schemes, adopting post-quantum algorithms such as lattice-based signatures, and introducing gradual upgrade mechanisms. It also presents a hypothetical scenario in which an attack on a blockchain forces an emergency hard fork.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

Audit finds 86% of tracked Algora bounties were unverified, cutting a bounty feed from 104 to about 33

A developer who runs a radar tracking paid GitHub issues audited the feed and found most of its flagged Algora bounties were not real. Of 97 issues marked as verified, 83 lacked any comment from the genuine Algora app, representing roughly $18,900 in claimed money. The errors stemmed from matching text like "/bounty $700" that anyone can type, look-alike bots mimicking Algora's format, and "AI only" issues posted by the same author. The developer now counts only issues confirmed by the real Algora account, which yields about 33 open issues today, and also checks that each is unassigned, unreferenced by merged pull requests, active, and not a crowded swarm. The author says the corrected, smaller numbers are more accurate and save effort, and is asking others which cheap verification signals they rely on.

0 comments Read more at DEV Community
0
ProgrammingDEV Community ·

nginx Buffering Delayed Claude Streaming Responses, Raising First-Token Time to Seconds

A developer found that their Claude chat app streamed smoothly locally but stalled for several seconds once deployed behind nginx. The cause was nginx's default proxy_buffering, which collects upstream data and sends it in chunks, turning token-by-token server-sent events into delayed bursts. Replaying 200 prompts showed median time to first token rising from 0.71 seconds to 3.4 seconds, with some replies arriving as a single block. Disabling proxy_buffering on the streaming route, or sending an X-Accel-Buffering: no header from the app, resolved the issue, though a regex location rule initially overrode the fix. After the change, median first-token time dropped to 0.76 seconds.

0 comments Read more at DEV Community