The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands in Step 6 using the {company} and {company-slug} variables, which are derived directly from the user-provided job description (JD). Because these variables are interpolated into bash commands (grep and ls) without sanitization, an attacker could provide a malicious JD containing shell metacharacters (e.g., ;, &&, `) to execute arbitrary commands through the Bash tool.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. It ingests untrusted job description content via $ARGUMENTS and uses this data to drive core logic (keyword extraction, tailoring decisions, and bullet point generation) without using boundary markers or instructions to ignore embedded commands.
Ingestion points: Job Description text or file path provided via $ARGUMENTS.
Boundary markers: None identified; external content is processed directly into the agent's context.
Capability inventory: File system access (Read, Write, Edit, Glob), shell execution (Bash), and local network communication (curl).
Sanitization: None; the skill is instructed to extract keywords and qualifications directly from the input text.

resume-tailor