Proactive For You recommendations
Built-in public sources provide initial video, article, research, podcast, audiobook, and visual choices without preference setup.
Product inventory
These statuses describe the public v0.6.2 source. Alpha means useful behavior exists, but contracts and storage may still change before 1.0.
Built-in public sources provide initial video, article, research, podcast, audiobook, and visual choices without preference setup.
Written direction plus topic and platform controls guide discovery while remaining optional.
More, Less, Hide, open, bookmark, topic, and source activity update a local recommendation profile.
Articles, images, feed-provided audio, and YouTube items open with recommendation evidence. Accessibility and media privacy hardening remain active work.
A briefing surface summarizes current discoveries. Its legacy and For You contracts are being consolidated.
Search local knowledge and use a reachable local Ollama model or an explicitly configured cloud provider for generated answers.
Publish video, images, audio, documents, slides, datasets, and other files with signed manifests and provenance evidence.
Direct peer HTTP transport, registry-assisted discovery, relay jobs, and encrypted messaging exist for controlled alpha use.
Invitations, friend-attributed recommendations, and intentional sharing among a small trusted group are the next social milestone.
Budgeted agent-to-agent services and broader untrusted-peer operation require additional protocol, safety, and abuse hardening.
Product tour
These are real product captures, not aspirational mockups. The interface continues to change during alpha.

Local node console. See daemon status, identity, content counts, and current recommendation highlights.
Available in alpha
For You. Guide attention, inspect evidence, open content, and teach the local profile through feedback.
Available in alphaImportant boundaries
RynMesh is alpha software. It is not yet a hardened open network for arbitrary untrusted peers, and its macOS community packages are not Apple-notarized.
Generated briefings and answers do require an available model.
Registry and peer infrastructure are optional for the first milestone.
Recommendation feedback stays in the node’s local state.