ccpi-marketplace

Warn

Audited by Gen Agent Trust Hub on Jul 27, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of the @intentsolutionsio/ccpi package from the public npm registry. This is a third-party tool that manages external code packs.
  • [REMOTE_CODE_EXECUTION]: The skill provides instructions to install and update plugin packs using ccpi install and built-in marketplace commands referencing jeremylongshore/claude-code-plugins. These commands download and execute external logic within the agent environment.
  • [COMMAND_EXECUTION]: The skill utilizes pnpm add -g for global package installation, which modifies the host environment and typically requires elevated permissions.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a high surface area for indirect injection as it ingests third-party plugin packs.
  • Ingestion points: External code is fetched during ccpi install or /plugin install operations.
  • Boundary markers: None identified; instructions within downloaded packs are not explicitly isolated from the agent's system prompt.
  • Capability inventory: The skill is granted Bash, Write, Read, and WebFetch tools, allowing downloaded packs to perform significant system operations.
  • Sanitization: There is no verification or sanitization mechanism described for the contents of the installed plugin packs.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 27, 2026, 11:38 AM
Security Audit — agent-trust-hub — ccpi-marketplace