The Agent Skills Directory

[COMMAND_EXECUTION]: The script 'cli.ts' is vulnerable to command injection. It takes the 'DEPTH' environment variable from the process environment and interpolates it directly into a shell command string executed by 'execSync'. If an attacker can control the 'DEPTH' variable (e.g., by tricking the agent into running the command with a crafted environment), they can execute arbitrary shell commands.
[DYNAMIC_EXECUTION]: The skill utilizes a dynamic code generation and execution pattern. In 'cli.ts', a Python script is stored as a string, written to a temporary file in the system's temporary directory at runtime, and then executed using the 'python3' interpreter.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it reads and displays untrusted data from the file system.
Ingestion points: The skill retrieves directory and file names using the 'fd' utility within the generated Python script.
Boundary markers: The output does not use delimiters or provide instructions to the agent to ignore potentially malicious text embedded in file names.
Capability inventory: The Bun script 'cli.ts' has capabilities for file system writes ('writeFileSync') and arbitrary command execution ('execSync').
Sanitization: No sanitization or escaping is performed on the file and directory names before they are printed to the console.

tree-view