skills/stophydotdev/skills/stophy-ads/Gen Agent Trust Hub

stophy-ads

Pass

Audited by Gen Agent Trust Hub on Oct 11, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill uses npx -y @stophy/cli *, which facilitates the download and execution of the vendor-owned CLI package from the npm registry. This is a standard method for distributing and running the vendor's tools.
  • [COMMAND_EXECUTION]: The skill executes shell commands using the stophy binary and npx to perform search operations and retrieve ad data. These commands are scoped to the primary functionality of searching public ad libraries.
  • [INDIRECT_PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by ingesting untrusted data from external ad platforms (Meta, Google, TikTok, LinkedIn).
  • Ingestion points: Search results and ad details retrieved from public ad libraries (documented in SKILL.md).
  • Boundary markers: The skill uses the --json flag to ensure output is returned in a structured format, reducing the risk of the agent misinterpreting ad copy as instructions.
  • Capability inventory: The skill utilizes Bash for command execution and file writing (via the -o flag).
  • Sanitization: External content is written directly to JSON files in the .stophy/ directory; no explicit prompt-level sanitization is mentioned, though structured data usage provides a natural boundary.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 11, 2026, 01:57 AM