OpenBot 0.38.0 simplifies remote sign-ins

// 1h agoPRODUCT UPDATE

OpenBot 0.38.0 simplifies remote sign-ins

OpenBot 0.38.0 lets users connect Codex, Claude, Gemini, and Muse to a joined remote server through their own computer’s browser. It also exposes host connectors like GitHub, Slack, Discord, and Telegram for remote management while keeping credentials on the server.

// ANALYSIS

This update tackles one of the least polished parts of self-hosted AI infrastructure: authentication across machines. OpenBot is making remote agent work feel more like a cohesive team platform than a collection of disconnected CLIs.

–Browser-based sign-in avoids forwarding provider credentials to remote clients
–Host-side token storage preserves the security boundary of self-hosted deployments
–GitHub, Slack, Discord, and Telegram connectors broaden OpenBot beyond coding workflows
–OAuth support for URL MCP servers makes third-party integrations easier to deploy
–The approach is especially useful for teams running agents on always-on servers or VPSs

// TAGS

openbotagentself-hostedmcpdevtoolopen-source

DISCOVERED

1h ago

2026-10-13

PUBLISHED

2h ago

2026-10-13

RELEVANCE

8/ 10

AUTHOR

OpenBot_

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE1h ago

Stalkr adds AI translation for global mentions

Stalkr now offers one-click AI translation for foreign-language mentions of tracked brands, products, and competitors. The update helps startup teams monitor global conversations without leaving their social-listening workflow.

UPDATE1h ago

Omarchy Teases Animated Theme Intros

Omarchy is bringing animated boot and theme intros to its opinionated Arch-based desktop, with community contributor @yamzeight showcasing the direction. Theme intro videos can play automatically at login and when switching themes.

OPEN SOURCE2h ago

QwenGyre open-sources long-horizon agent RL

Alibaba’s Qwen team has open-sourced QwenGyre, an online RL framework for training agents through tasks that can run for hours, span hundreds of interactions, and consume nearly 1M tokens. It elastically reallocates GPUs between rollout and training, reporting up to 1.85× speedups and a six-point NL2RepoBench gain on Qwen3.8 2.4T.