Searches past AI sessions and messages in Taplex data using SQLite's built-in FTS5 text indexing.
npx skills add https://github.com/juliuschun/tower --skill search-juliuschun
Please help me install this Agent Skill. Skill: search Source: https://github.com/juliuschun/tower/tree/main/claude-skills/skills/search Command: npx skills add https://github.com/juliuschun/tower --skill search-juliuschun
This Skill addresses the challenge of lost or hard-to-find information within past AI conversations and sessions, enabling users to quickly retrieve valuable context and insights.
tower.db database./search api integration to find the relevant session and its key discussion points.Search for past conversations about 'database optimization' by typing /search database optimization.
High-intent search queries and answers about installing and using this skill.
You can search past AI conversations by running a keyword query against the tower.db database. This Skill leverages FTS5 full-text search to instantly retrieve specific discussion points, session summaries, and historical context from past sessions.
Yes, you can find specific technical decisions by searching session names and message content. For example, querying a specific phrase like API integration retrieves the relevant past session along with its key discussion points and summaries.
Yes, the tool provides database statistics and recent session timelines. It offers insights into database usage, including total session counts, costs, and model distribution, alongside a timeline of recent AI sessions.
The best way to analyze past AI session histories and costs is to use the session analysis feature. It allows users to view summaries, associated costs, and complete message histories of specific past sessions stored in the database.
FTS5 full-text search works by indexing session names and message content within the tower.db database. This mechanism enables efficient retrieval of relevant conversation snippets and summaries based on your input keywords.
PedroGiudiceSearch past session observations stored in a local DuckDB memory store.
Owl-ListenerDesign AI context window composition with structured budgeting and ordering.
ericwang915Aggregate token usage and API costs from the usage history log.
sixscripts-aiConfigure vector retrieval pipelines with embedding models and HNSW indexes.