codebase-adapter

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses Bash to run ls, cat, python3, and node for project discovery. These commands are used throughout the workflow to verify the file system, read configuration files, and identify the technology stack.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it reads data from potentially untrusted project files and interpolates it into the agent's core instructions.
  • Ingestion points: Project files such as package.json (scripts and dependencies) and .claude/settings.json (MCP server names), along with file system paths.
  • Boundary markers: No delimiters or warnings are used when incorporating external data into the agent's markdown configuration files.
  • Capability inventory: The skill has permissions to execute shell commands (ls, cat, node, python3) and performs file modifications.
  • Sanitization: There is no evidence of validation or sanitization for the data read from the host project before it is used to modify the plugin's skills and agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 06:11 PM