ddg-search-hardened

Pass

Audited by Gen Agent Trust Hub on Apr 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script (scripts/search.py) to retrieve search results from DuckDuckGo.
  • [EXTERNAL_DOWNLOADS]: The skill requires the ddgs Python package, which is a standard library for the intended search functionality.
  • [PROMPT_INJECTION]: The skill processes untrusted web content from search results, which constitutes an indirect prompt injection surface.
  • Ingestion points: Search result titles, snippets, and URLs are retrieved and displayed by scripts/search.py.
  • Boundary markers: The SKILL.md file contains explicit 'Security Guardrails' instructing the agent to treat search results as untrusted and to ignore any instructions embedded within them.
  • Capability inventory: The skill uses shell execution for the search script and references the use of web_fetch for further content retrieval.
  • Sanitization: No programmatic sanitization of search result text is performed within the script itself.
  • [SAFE]: The skill includes comprehensive documentation (SAFETY.md) and specific instructions in SKILL.md to mitigate risks associated with SSRF and unauthorized data transmission to external endpoints.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 21, 2026, 06:31 PM
Security Audit — agent-trust-hub — ddg-search-hardened