optimize-plugin

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute a local Python validation script (validate-plugin.py) located in the environment's plugin root directory. This involves executing shell commands with a user-provided path variable ($TARGET), which presents a command execution surface.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection.
  • Ingestion points: Untrusted data enters the agent context through the provided by the user, specifically when the skill reads the .claude-plugin/plugin.json manifest and other component files in the commands, agents, and skills directories (SKILL.md).
  • Boundary markers: The skill lacks explicit boundary markers or 'ignore embedded instructions' warnings when passing the content of the target plugin and the validation results to the plugin-optimizer agent in Phase 2 (SKILL.md).
  • Capability inventory: Across its phases, the skill employs Bash(bash:*), Task (to launch secondary agents), Read, and Glob tools (SKILL.md).
  • Sanitization: There is no evidence of sanitization, escaping, or validation of the external plugin content before it is interpolated into the instructions for the optimization agent (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 10:16 AM