printing-press-score

Warn

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using identifiers and paths derived from free-form user input. Specifically, the <resolved-path> and <spec-path> variables are interpolated into printing-press scorecard commands executed via the Bash tool. Without strict sanitization of these variables, an attacker could provide input containing shell metacharacters (e.g., ;, "", $()`) to achieve arbitrary command execution.
  • [COMMAND_EXECUTION]: The setup contract in SKILL.md includes logic to prepend the current workspace or git top-level directory to the PATH if an executable named printing-press is found there. This behavior allows for potential binary hijacking or execution of untrusted local code if the agent is operating within a directory containing malicious files.
  • [EXTERNAL_DOWNLOADS]: The instructions recommend installing the printing-press utility from the author's GitHub repository (github.com/mvanhorn/cli-printing-press). This is an expected dependency for the skill's core functionality and originates from the same vendor context as the skill itself.
Audit Metadata
Risk Level
MEDIUM
Analyzed
May 14, 2026, 07:19 PM
Security Audit — agent-trust-hub — printing-press-score