clipboard-file

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The test fixture fixtures/agentic_eval.json references a Python script located outside the skill's root directory (../../best-practices-skills/scripts/validate_skill.py). This creates a dependency on a specific external file structure for testing.\n- [COMMAND_EXECUTION]: The bundled shell script scripts/copy-file-to-clipboard.sh executes several system utilities (xclip, pgrep, ps, kill) and launches background processes using setsid. These operations are required for managing the clipboard lifecycle.\n- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection via user-supplied file paths.\n
  • Ingestion points: The file_path argument processed in SKILL.md and scripts/copy-file-to-clipboard.sh.\n
  • Boundary markers: None identified.\n
  • Capability inventory: Execution of xclip, pgrep, ps, and kill via shell commands in scripts/copy-file-to-clipboard.sh.\n
  • Sanitization: Employs realpath -e for path validation and passes variables as positional parameters to sh -c to mitigate basic shell injection.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 06:00 PM
Security Audit — agent-trust-hub — clipboard-file