Menu

#Capability

19 posts

Feed·
19 of 19 posts
How to manage multi-user AI agent authentication and authorization in 2026 (OAuth 2.1, OIDC, and delegated access)
🖼️
0

How to manage multi-user AI agent authentication and authorization in 2026 (OAuth 2.1, OIDC, and delegated access)

DEV Community·Manveer Chawla·18 days ago
#4ilaX9Ei
#capability#pattern#how#agent#user#runtime

From Dev.to - security: How to manage multi-user AI agent authentication and authorization in 2026 (OAuth 2.1, OIDC, and delegated access)

15s
Read More
I Built an Offline AI Compliance Auditor for Indian Factories Using Gemma 4 — Here's Everything I Learned
🖼️
0

I Built an Offline AI Compliance Auditor for Indian Factories Using Gemma 4 — Here's Everything I Learned

DEV Community·Simran Shaikh·20 days ago
#gRzQqHDF

A complete guide covering local setup, model comparison, multimodal deep dive, fine-tuning, and why a model this capable running locally changes everything — from an Indian developer's perspective.

15s
Read More
Security Architecture Proposal: Mitigating Supply Chain Vulnerabilities via Linear Capability-Based Permissions
🖼️
0

Security Architecture Proposal: Mitigating Supply Chain Vulnerabilities via Linear Capability-Based Permissions

DEV Community·suissAI·26 days ago
#WlWWskWj
#austral#security#suissa#aurora#linear#system

From Dev.to - security: Security Architecture Proposal: Mitigating Supply Chain Vulnerabilities via Linear Capability-Based Permissions

15s
Read More
Mythos Found a 27-Year-Old Bug in OpenBSD. Your Code Is Next.
🖼️
0

Mythos Found a 27-Year-Old Bug in OpenBSD. Your Code Is Next.

DEV Community·Michelle Jones·29 days ago
#GVYzi4dW
#security#ai#legacy#mythos#model#capability

Anthropic's new Mythos model surfaced a 27-year-old vulnerability in OpenBSD and generated 181 Firefox exploits where Opus 4.6 produced two. Eleven organizations have launch access. The rest of us have a window.

15s
Read More
The lethal trifecta in two-agent practice: seven incidents in 48 hours
🖼️
0

The lethal trifecta in two-agent practice: seven incidents in 48 hours

DEV Community·Dutch AI Agents·29 days ago
#ePk8BduA
#leg#ai#security#capability#agent#farcaster

Field answer to Simon Willison's 'lethal trifecta' question, written from inside a two-agent autonomous setup with private keys, untrusted text, and unrestricted external channels. Seven documented incidents across all three legs in 48 hours.

15s
Read More
Wrap the GitHub MCP Registry as a Naftiko Capability — REST, MCP, and Agent Skills From One Spec
🖼️
0

Wrap the GitHub MCP Registry as a Naftiko Capability — REST, MCP, and Agent Skills From One Spec

DEV Community·Kin Lane·about 1 month ago
#mXB6QSUH
#three#why#mcp#registry#naftiko#capability

From Dev.to - api: Wrap the GitHub MCP Registry as a Naftiko Capability — REST, MCP, and Agent Skills From One Spec

15s
Read More
Coffee Break: Armed Madhouse - The U.S. Navy Adrift | naked capitalism
🖼️
0

Coffee Break: Armed Madhouse - The U.S. Navy Adrift | naked capitalism

naked capitalism·Haig Hovaness·about 1 month ago
#49EvHJ7i
#comment#respond#masthead#force#navy#naval

America’s naval challenges are not isolated breakdowns but interacting structural failures. From procurement dysfunction and support-vessel attrition to combat lessons from Iran and the growing mission-capability gap, U.S.…

15s
Read More
Where Capability Lives: A Meta-Protocol for Distributed Intelligence on the Trillion-Device Installed Base
📰
0

Where Capability Lives: A Meta-Protocol for Distributed Intelligence on the Trillion-Device Installed Base

DEV Community·Rotifer Protocol·about 1 month ago
#iEUgNKxZ
#ai#evolution#protocol#capability#layer#meta

From Dev.to - opensource: Where Capability Lives: A Meta-Protocol for Distributed Intelligence on the Trillion-Device Installed Base

15s
Read More
📰
0

The Dual-Use Problem Is a Trust-Architecture Problem

DEV Community: security·Alex @ Vibe Agent Making·about 1 month ago
#0VJirixA

In January 2026, a seventeen-year-old remote code execution vulnerability sat undiscovered in FreeBSD's NFS implementation. CVE-2026-4747 required chaining six sequential RPC requests through a stack buffer overflow in the RPCSEC_GSS authentication…

15s
Read More