risk-management

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes a local Python script scripts/risk_management_diagnostics.py to perform data analysis. This script is distributed within the skill package and operates deterministically on local CSV data.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data from input.csv. While the included script focuses on numeric aggregation which naturally limits injection risks, the underlying ingestion pattern represents a standard surface for indirect prompt injection if the raw data is later processed by the LLM.
  • Ingestion points: Data is read from input.csv as specified in the SKILL.md workflow.
  • Boundary markers: No explicit delimiters or instruction-isolation markers are present in the skill's prompts to handle external data.
  • Capability inventory: The skill possesses the capability to execute local scripts via the command line to process data.
  • Sanitization: The diagnostic script performs numeric type coercion and data cleanup using pandas, effectively sanitizing input intended for mathematical summarization.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 08:46 AM
Security Audit — agent-trust-hub — risk-management