The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on shell commands to perform its primary function. It instructs the agent to use standard utilities such as find, grep, wc, awk, sed, and xargs to analyze file lengths, count specific code patterns (like React hooks or database queries), and identify potentially duplicated or legacy files.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it ingests data from the project codebase, including file paths and contents, which are then processed by shell commands and the agent's logic.
Ingestion points: The skill reads directory structures and file contents from the local file system via find and grep commands.
Boundary markers: No specific boundary markers or instructions to ignore embedded malicious prompts within the audited files are present.
Capability inventory: The agent has the capability to execute shell commands and read/write files as part of the auditing process (e.g., find, grep, xargs, wc).
Sanitization: There is no evidence of sanitization for file paths or file content before they are used as arguments in shell commands or analyzed by the agent.

common-architecture-audit