kai-case-study
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts in Phase 4 (
scripts/quality_gates/four_us_score.pyandscripts/quality_gates/banned_word_check.py) to perform quality gates on generated content. These scripts are invoked via the command line with generated file paths as arguments. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources during the discovery phase which represents a potential injection surface.
- Ingestion points: Reads interview transcripts and survey responses provided by the user in Phase 1.
- Boundary markers: The instructions do not specify the use of delimiters or 'ignore' instructions when processing the source material.
- Capability inventory: The agent can execute local Python scripts (Phase 4) and write output files to the
workspace/directory (Phase 5). - Sanitization: There is no mention of sanitization or validation of the source material before it is incorporated into the prompt context for case study generation.
Audit Metadata