thinkt indexer#

Specialized indexing and search via DuckDB (requires thinkt-indexer)

Synopsis#

The indexer command provides access to DuckDB-powered indexing and search capabilities. This requires the ’thinkt-indexer’ binary to be installed separately due to its CGO dependencies.

Examples: thinkt indexer start # Start indexer in background thinkt indexer status # Check indexer status thinkt indexer stop # Stop background indexer thinkt indexer sync # Sync all local sessions to the index thinkt indexer search “query” # Search across all sessions thinkt indexer summarize tags “trace tag” # Suggest shareable tags

Options#

      --db string              path to DuckDB index database file
      --embeddings-db string   path to DuckDB embeddings database file
  -h, --help                   help for indexer
      --log string             path to log file
  -q, --quiet                  suppress progress output
  -v, --verbose                verbose output

SEE ALSO#

• thinkt - Tools for AI assistant session exploration and extraction
• thinkt indexer embeddings - Manage embedding model, storage, and sync
• thinkt indexer help - Help about any command
• thinkt indexer logs - View indexer logs
• thinkt indexer metrics - Show Prometheus metrics from the running indexer
• thinkt indexer search - Search for text across indexed sessions
• thinkt indexer semantic - Semantic search and index management
• thinkt indexer sessions - List sessions for a project from the index
• thinkt indexer start - Start indexer in background
• thinkt indexer stats - Show usage statistics from the index
• thinkt indexer status - Show indexer status
• thinkt indexer stop - Stop background indexer
• thinkt indexer summarize - Manage summarization model, storage, sync, and tags
• thinkt indexer sync - Synchronize all local sessions into the index
• thinkt indexer version - Print version information