luci-memoryv1.0.2

Users can search their personal video library using natural language queries like 'what did I do last weekend' or 'show me videos from my trip to Paris'. The skill retrieves relevant videos, keyframes, and transcripts, converting UTC timestamps to local time.

When a user wants to recall conversations or events involving specific people, this skill searches portrait data including traits, speeches, and relationships. For unknown individuals, it falls back to transcript search for their name.

Content creators can locate specific moments in their video archives using queries about what was said or shown. The skill supports multi-step retrieval: first locating videos, then fetching complete transcripts via audio or visual searches.

Users can explore their own personality traits, hobbies, and relationship patterns extracted from their video data. Queries like 'what are my hobbies?' or 'who are my close friends?' are answered by searching portrait data.

To answer 'what happened at the team meeting last week', the skill combines event searches, video retrieval, and transcript analysis to reconstruct a timeline. It ensures no hallucination by grounding every claim in retrieved data.

Offer tiered monthly subscriptions for personal video memory search and analysis. Basic tier includes limited searches, premium tier offers unlimited queries and advanced portrait insights. Recurring revenue from individual users.

License the skill to enterprises for team knowledge management. Teams can search meeting recordings, retrieve action items, and analyze communication patterns. Revenue from per-seat licensing and custom integrations.

Provide API access to developers who want to integrate memory search into their own applications (e.g., CRM, note-taking apps). Charge per API call or monthly usage tiers. Additional revenue from premium support.