notemdpro

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection attack surface as it is designed to process untrusted data from web research and existing user notes.
  • Ingestion points: Processes markdown files from the local vault and external content from web research services.
  • Boundary markers: No explicit delimiters or safety instructions are defined to separate untrusted data from agent instructions.
  • Capability inventory: Includes file system write access and network communication capabilities.
  • Sanitization: Lacks documented sanitization or validation of input data.
  • [DATA_EXFILTRATION]: The skill's research functionality communicates with external domains such as Tavily and DuckDuckGo. While these are intended functions, they involve network activity to domains outside the standard whitelist.
  • [EXTERNAL_DOWNLOADS]: Documentation points to external GitHub repositories for installation and setup. These references target a well-known service for legitimate development purposes.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 02:49 PM