israeli-bagrut-psychometric
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied academic data such as subject names, units, and grades to perform score calculations. This represents a standard indirect prompt injection surface. Ingestion points: The skill ingest user input through the agent prompt when providing grade and subject information. Boundary markers: None explicitly defined within the instructions; relies on standard model parsing. Capability inventory: The skill uses a local script scripts/bagrut-calculator.py for arithmetic operations. It lacks network access or file-writing capabilities. Sanitization: The calculation script performs strict type conversion (float/int) on input data, which effectively prevents the execution of arbitrary command strings via the calculation parameters.
- [COMMAND_EXECUTION]: The skill instructions guide the agent to run a local Python script to perform mathematical calculations. Evidence: Examples in SKILL.md and SKILL_HE.md show command patterns like 'python scripts/bagrut-calculator.py --mode subject --exam 82 --magen 90'. This is a legitimate and expected use of a local tool for the skill's primary purpose.
Audit Metadata