lean-target

Pass

Audited by Gen Agent Trust Hub on Nov 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes git clone to fetch external repositories and runs "build/test commands" discovered from the target's files and evidence (SKILL.md, Step 4).
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to read and follow "the target's own instructions" (SKILL.md, Step 1) and "discover build/test commands from target evidence" (SKILL.md, Step 4). This allows malicious instructions or commands inside a cloned repository to influence the agent's behavior.
  • Ingestion points: Remote repository URLs provided by the user and files (instructions, evidence) within the cloned repository (SKILL.md).
  • Boundary markers: The skill lacks explicit boundary markers or instructions to ignore embedded prompts when reading target repository content.
  • Capability inventory: Performs shell command execution via git clone and arbitrary discovered build/test commands (SKILL.md).
  • Sanitization: The skill includes some safeguards, such as rejecting credentials embedded in URLs and requiring paths to be passed as literal arguments to prevent shell interpolation (SKILL.md, Step 1).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 10, 2026, 02:14 PM