skills/quantskills/skill-index-rebalance-event-study/index-rebalance-event-study/Gen Agent Trust Hub
index-rebalance-event-study
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The validation script (validation/smoke.py) and unit tests (tests/test_index_rebalance.py) utilize subprocess.run to execute the local analysis script scripts/study_index_rebalance.py for verification. These calls use static paths within the skill directory.
- [DYNAMIC_EXECUTION]: The test suite (tests/test_index_rebalance.py) uses importlib.util to dynamically load the scripts/study_index_rebalance.py module. This allows for unit testing of internal functions without a complex package setup and is restricted to the local file.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted CSV data for analysis. Evidence chain: 1. Ingestion points: scripts/study_index_rebalance.py via the --input CLI argument. 2. Boundary markers: Not applicable as data is processed via CSV parser. 3. Capability inventory: Arithmetic return calculations and JSON report generation. 4. Sanitization: The script performs rigorous type validation, checking that numeric fields are finite numbers and dates match the YYYY-MM-DD format.
Audit Metadata