agent-browser supports safer TLS proxies

// 1d agoPRODUCT UPDATE

agent-browser supports safer TLS proxies

agent-browser’s Linux Chromium workflow accepts a proxy CA certificate through CLI, environment, config, or MCP settings. It enables HTTPS interception without disabling normal hostname and certificate validity checks.

// ANALYSIS

This is a meaningful security upgrade for browser agents operating inside corporate networks, where broad certificate bypasses are an unacceptable shortcut.

–Isolated NSS trust limits the trusted CA to the launched Chromium session
–`--ca-cert` avoids the security risks of `--ignore-https-errors`
–Retained trust across commands simplifies long-running agent workflows
–Proxy CA support targets Zscaler, Netskope, mitmproxy, and similar TLS-intercepting environments
–Linux-only support and incompatibility with CDP, profiles, and cloud providers remain notable constraints

// TAGS

agent-browserdevtoolcliweb-agentsecurityopen-source

DISCOVERED

1d ago

2026-08-29

PUBLISHED

1d ago

2026-08-28

RELEVANCE

8/ 10

AUTHOR

ctatedev

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

MODEL43m ago

FastH3 V1 drops 14x-faster MiniMax video

FastVideo has open-sourced FastH3 Preview v1, a four-step DMD2-distilled MiniMax H3 model for text-to-video-and-audio with 90% sparse attention. The open weights and LoRA target up to 14× faster Blackwell inference, with 15-second 768p clips generated in under 13 seconds across eight B200 GPUs. [Announcement](https://haoailab.com/blogs/fasth3-preview/) [Model card](https://huggingface.co/FastVideo/FastVideo-FastH3-4-step-Preview-v1-LoRA)

MODEL43m ago

fal H3 Max makes video generation real-time

fal’s H3 Max is a post-trained, inference-optimized variant of MiniMax H3, tuned for prompt adherence and visual quality. It generates five-second videos in under three seconds through fal’s API.

OPEN SOURCE43m ago

VoiceMem brings streaming memory to voice agents

VoiceMem is an Apache-2.0 memory layer for voice agents that separates factual recall from emotional and persona context. Its streaming architecture retrieves a small set of relevant memories while users are still speaking, targeting personalization without added conversational delay.