datetime

Pass

Audited by Gen Agent Trust Hub on Feb 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill directs the agent to use the Bash tool for date calculations and natural language parsing using the date -d command. It includes examples of shell variable assignment and arithmetic expansion, such as calculating days between dates using Unix timestamps. It also references the execution of a local script located at ~/.claude/skills/week-mapping.sh for academic week calculations.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by instructing the agent to pass untrusted user input directly into shell commands.
  • Evidence Chain for Indirect Injection:
  • Ingestion points: Natural language temporal expressions provided by the user (e.g., 'tomorrow', 'next week') processed in SKILL.md.
  • Boundary markers: Absent; expressions are directly embedded into the date -d command string.
  • Capability inventory: Subprocess execution via the Bash tool with access to shell features like command substitution and local file paths.
  • Sanitization: Absent; there are no instructions for the agent to validate, escape, or sanitize the user input to prevent the execution of shell metacharacters such as $(...), backticks, or command separators.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 28, 2026, 11:04 PM