skills/klh/speedy-claude/find-skills/Gen Agent Trust Hub

find-skills

Warn

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes shell commands via npx skills to interact with a package ecosystem, performing keyword searches and managing skill installations.
  • [EXTERNAL_DOWNLOADS]: Facilitates downloading agent skills from remote sources, specifically targeting GitHub repositories and the skills.sh registry.
  • [REMOTE_CODE_EXECUTION]: Explicitly instructs the agent to use the -y flag when adding skills (npx skills add <package> -y). This flag bypasses interactive confirmation prompts, allowing for the silent installation and subsequent execution of third-party code.
  • [PROMPT_INJECTION]: The skill's workflow for finding and installing external packages creates a surface for indirect prompt injection, where malicious package metadata could influence agent actions.
  • Ingestion points: User search queries processed via npx skills find.
  • Boundary markers: None identified.
  • Capability inventory: Subprocess execution for package discovery (npx skills find) and installation (npx skills add).
  • Sanitization: No input validation or sanitization of search queries or package names is specified.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 24, 2026, 12:49 PM
Security Audit — agent-trust-hub — find-skills