find-skills

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the agent to install the skills CLI via npm and uses npx to dynamically run the @nacos-group/cli package for registry interaction.
  • [COMMAND_EXECUTION]: The agentteams-find-skill.sh script executes various system commands, including npm for global package installation, npx for dynamic tool execution, and curl for interacting with the registry API and obtaining authentication credentials.
  • [REMOTE_CODE_EXECUTION]: The skill is designed to fetch and install third-party 'skills' from external registries. These skills are modular packages that extend the agent's capabilities with new code and workflows.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data in the form of skill descriptions and metadata from external registries.
  • Ingestion points: Search results and skill listings are retrieved from skills.sh and Nacos-based registries (scripts/agentteams-find-skill.sh).
  • Boundary markers: The skill does not implement specific delimiters to isolate registry content from the agent's core instructions.
  • Capability inventory: The skill environment allows for shell command execution, network requests, and the installation of executable code packages.
  • Sanitization: The search query processing includes alphanumeric filtering to prevent command injection, though the content of the installed skills themselves is not inspected by this tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 02:47 PM