Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
Detects clone farming in AI agent marketplaces by analyzing content similarity, batch publish patterns, ID washing, and cross-citation rings across skills. Takes a list of skills, a publisher node ID, or a search term as input and returns a structured risk report with cluster groups and a CLEAN/SUSPECT/FARMING rating.
Most marketplaces only deduplicate by exact hash, so this catches ID-washed near-duplicates and coordinated publish patterns that standard checks miss.
When to use it
- Auditing top search results before integrating a marketplace skill
- Checking a publisher's catalog for batch-published near-duplicates
- Identifying ID-washed copies that bypassed hash deduplication
- Spotting cross-citation rings inflating dependency trust scores
- Investigating a sudden flood of similarly named skills from one node
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Scan top 10 results for 'code formatter' on the marketplace
- 1Fetches top 10 search results with their Capsule source and metadata
- 2Computes pairwise similarity scores across all skill pairs
- 3Checks publish timestamps for batch patterns from the same node
- 4Traces dependency declarations to find cross-citation rings
- 5Generates a risk-rated report grouping suspects into clusters
Report showing 6 of 10 results are clones from one publisher across 2 clusters, with similarity scores, publish timeline, ID washing evidence, and a recommendation to flag the publisher for review