Actions

1. ExtractTextFromPDF

Parameters: pdfBinary, normalizeWhitespace, collapseRepeatedNewlines, includePageNumberPrefix, maxTotalChars, collectLogs, attachFilesLogs.\

Output: Text + logsZipFile.

2. ChunkPlainText

Parameters: text, chunkSizeChars, overlapSizeChars, normalizeWhitespace, estimateTokens, maxTotalChars, collectLogs, attachFilesLogs.\

Output: Chunk[] + stats + logsZipFile.

3. ExtractAndChunkPDF

One-step pipeline (extraction + chunking). Returns Chunk[] + stats + logsZipFile.

Add the ExtractAndChunkPDF Action in a client action or a workflow, then use the Chunk Output in a loop to create embeddings and store in your VectorDB