DeclawedScan AI Skills & Prompts
for Malicious Content
Multi-engine threat detection for AI skills, system prompts, MCP servers, and agent configurations. Protect against prompt injection, data exfiltration, and tool poisoning.
By submitting data above, you are agreeing to our Terms of Service and Privacy Notice, and to the sharing of your submission with the security community.
Detection Engines
Multi-Engine Detection
6 specialized engines analyze content: PatternGuard, YARAScanner, HeuristicScan, FlowTracer, LLMAnalyzer, and CommunityRules.
Instant Results
Get comprehensive security analysis in seconds. Cached results for known hashes are returned in milliseconds.
Community Intelligence
Benefit from crowdsourced threat intelligence. Every scan contributes to our growing database of known threats.
What We Detect
Attempts to override system instructions
Sending sensitive data externally
Targeting API keys, tokens, passwords
Targeting wallet keys and seed phrases
Malicious tool descriptions
Safety measure bypasses
Unicode tricks, invisible content
Encoded or encrypted payloads