One command. Your agent gets a video backend.

Rimsha Bhardwaj @heyrimsha

BREAKING: Someone just built the open source Loom killer that turns your screen recordings into AI agents. It's called Bloom and it does something Loom never could. Your recordings don't just sit as files. They become queryable, searchable, agent-ready data the moment you stop recording. You hit record. Bloom captures your screen, mic, and system audio locally. While you're still recording, it uploads chunks to the cloud in real time. The second you stop, it auto-generates a full transcript, creates visual embeddings, and indexes every spoken word. You can query your recordings through APIs and hook Claude Code directly into your video library. Recordings are no longer files. They're inputs for AI. MIT License. 100% Opensource.

Erick @ErickSky

¿Y si tu próxima reunión tuviera un AI coach susurrándote al oído en tiempo real? Imagina esto: estás en una reu importante… grabando TODO localmente (nada se va a la nube), transcripción en tiempo real separando tu voz de la del cliente, y un agente IA te avisa mientras hablas: «estás hablando de más», «haz esta pregunta ahora», «el cliente está dudando aquí». Y cuando cuelgas… resumen perfecto, action items automáticos, export directo a Notion / CRM. Se llama call.md y es open-source + 100% local-first. Esto no es otro notetaker. Es un agente vivo dentro de tu llamada. REPOOO

Vish @rv_RAJvishnu

Screen awareness is huge because most real debugging work happens outside the code file itself. logs, browser state, and tiny UI regressions are where hours disappear... once an agent can watch that context directly, prompting gets simpler and the loop gets much faster. if this ships with solid privacy controls, it could become default workflow for people who build all day.

Aman Hanspal @han__ama

One problem I kept thinking about while building a side project: "how do you actually make events in videos searchable?" Turns out @videodb_io is working on exactly this. Attended their session in Mumbai recently — interesting to see how they're approaching video querying at a systems level.

Sushant Rai Building AI sous chef

Recently attended "Give your AI agents eyes and ears" workshop by VideoDB. What stood out most was their approach: rather than forcing the AI to scrub through footage linearly, they rely on real-time indexing and semantic search.

BombasticSideEye @BombasticEye77

We just hit the agents watching content for you era. Summarising, tagging, answering questions… Coral's videodb director is basically giving agents eyes and contextual memory. Wild upgrade.

Twlv @twlvone

The screen context is what changes everything. half the time claude code fails its because it cant see what you see. adding that visual loop closes the gap between 'impressive demo' and 'actually useful daily tool'.

Arthur Anderson @Arthur_Ande

Should've had the VideoDB Director Agent in my university days to grab quick summaries to learn easily. AI Agents are getting a lot smarter through @Coral_Protocol.