skill-forge-benchmark

Pass

Audited by Gen Agent Trust Hub on Nov 10, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes a local Python script scripts/aggregate_benchmark.py to aggregate performance metrics. This is a core part of its functionality but involves shell command execution.
  • [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests and processes potentially untrusted data from external evaluation sets and results.
  • Ingestion points: The skill reads data from evals/evals.json, timing.json, and grading.json files.
  • Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are specified for the processed data.
  • Capability inventory: The skill possesses the ability to execute shell commands (Python scripts) and read/write files in the workspace.
  • Sanitization: The instructions do not define any sanitization or validation steps for the content of the evaluation data before processing it through the aggregation script or report generation.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 10, 2026, 04:45 PM