NYC

image-search

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION] (LOW): The skill requests access to the Bash tool to execute a local script scripts/image_search.py. While standard for this functionality, it utilizes a high-privilege capability.
  • [EXTERNAL_DOWNLOADS] (LOW): The skill connects to DuckDuckGo for searches and downloads image files from external domains.
  • [PROMPT_INJECTION] (LOW): The skill is susceptible to Indirect Prompt Injection (Category 8). 1. Ingestion points: Untrusted data from web search results (titles, source URLs, and metadata) is ingested via the search script output. 2. Boundary markers: No delimiters or safety instructions are present to prevent the agent from interpreting search metadata as commands. 3. Capability inventory: The skill has access to Bash, Write, and Edit tools, which could be exploited if malicious content is processed. 4. Sanitization: No sanitization or validation of the ingested external data is mentioned in the provided instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:19 PM