The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to dynamically construct and execute shell commands using the $ARGUMENTS variable (e.g., cargo clippy -p <package>). Because there are no instructions to sanitize or validate this user-provided input, it is vulnerable to command injection if a user provides input containing shell metacharacters (e.g., ;, &, |).
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8) because it parses output from external tools (cargo clippy) and uses that data to drive its 'Decision Framework' for editing files.
Ingestion points: Output from cargo clippy execution in SKILL.md (Step 1 and Step 3).
Boundary markers: Absent; the agent is instructed to 'categorize each remaining warning' directly from the tool's stdout/stderr.
Capability inventory: The skill is authorized to use Bash, Read, and Edit tools, allowing it to modify source code based on what it perceives in the warning messages.
Sanitization: Absent; there is no validation of the warning content to ensure it does not contain malicious instructions or misleading filenames designed to trick the agent into performing unauthorized edits.

rust-warn-fix