Cases Skills About

experiments-index - Incrementally append records to FAISS and HuggingFace store

Appends campaign result records to an existing FAISS index and HuggingFace dataset under an fcntl file lock.

Tags

Updated: 2026-09-18

faiss huggingface-datasets vector-store incremental-append rag

Capabilities

Acquire exclusive fcntl file lock Embed text records with encoder Normalize embeddings using L2 norm Append vectors to FAISS index

Typical Inputs

Campaign result records Target FAISS index file path Target HuggingFace dataset directory path

Typical Outputs

Updated FAISS index file Updated HuggingFace Arrow dataset directory JSON summary of appended count and total index size

What this skill does

Acquire exclusive fcntl file lock
Embed text records with encoder
Normalize embeddings using L2 norm
Append vectors to FAISS index
Concatenate records into Arrow dataset
Save updated store to disk

Inputs

Campaign result records
Target FAISS index file path
Target HuggingFace dataset directory path
Sentence-embedding encoder model

Outputs

Updated FAISS index file
Updated HuggingFace Arrow dataset directory
JSON summary of appended count and total index size

Requirements

Python environment with faiss-cpu or faiss
Python HuggingFace datasets package
Sentence-embedding encoder model
File system supporting fcntl locks

Source

Spec: SKILL.md

ClawIndex

OpenClaw Skills & Use Case Index

ClawIndex is an ecosystem-driven index of OpenClaw skills and real-world use cases.

Index

Meta

© 2026 ClawIndex All Rights Reserved.