score
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a shell command (
node scripts/score.mjs --findings <path>) where the file path is sourced directly from user input ($ARGUMENTS). This interpolation creates a risk of command injection if the platform does not sanitize the user-provided arguments before execution.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external JSON findings files, creating an attack surface for indirect prompt injection.\n - Ingestion points: User-provided findings JSON file via
$ARGUMENTS(SKILL.md).\n - Boundary markers: None are defined to separate untrusted JSON data from the agent's instructions.\n
- Capability inventory: The skill uses
Bashto execute processing scripts andReadto access local files.\n - Sanitization: There is no evidence of content validation or sanitization for the JSON data before it is re-calculated and presented back to the agent context.
Audit Metadata