generate-test-data

Pass

Audited by Gen Agent Trust Hub on Oct 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill requires running TypeScript/ESM scripts (inspect-workbook.mts and generate-fixtures.mts) using Node.js to parse workbooks and produce output. These commands are intended for local development workflows.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of Excel workbooks (.xlsx, .xls).
  • Ingestion points: The fileProcessor.load([XLSX_FILE]) call in scripts/generate-fixtures.mts and scripts/inspect-workbook.mts reads the workbook content.
  • Boundary markers: The scripts parse specific decision table schemas; however, there are no explicit 'ignore instructions' markers for the content within cells.
  • Capability inventory: The skill can write files (fs.writeFile), create directories (fs.mkdir), and delete files (fs.rm) within the designated OUT_DIR.
  • Sanitization: The output is generated via JSON.stringify, ensuring the resulting fixtures are valid JSON, though it does not sanitize the logical content of the strings for downstream LLM consumers.
  • [DATA_EXPOSURE_AND_EXFILTRATION]: The FixtureWriter class in scripts/generate-fixtures.mts performs file system operations to manage test fixtures. It includes a cleanup step that removes existing .json files in the output directory before generating new ones. These operations are scoped to the local filesystem and do not involve network transmission.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 4, 2026, 12:37 PM