AISPA: User-Centric System Prompt Auditing for Large Language Model Applications
AISPA turns system-prompt review into an eight-dimension audit. Its survey suggests builders should test prompts for user protection and conflicting instructions, not merely check that safeguards exist.
AISPA classifies **3,249 instructions** from **88 commercial AI products** as protective or problematic across eight user-centered dimensions. Although 98.9% of products include a protection, only 24% cover every dimension.
Audit agent system prompts instruction by instruction, checking both coverage and conflicts. A long safety section is weak evidence when protective and user-hostile directives can coexist in the same prompt.
AISPA classifies **3,249 instructions** from **88 commercial AI products** as protective or problematic across eight user-centered dimensions. Although 98.9% of products include a protection, only 24% cover every dimension. Audit agent system prompts instruction by instruction, checking both coverage and conflicts. A long safety section is weak evidence when protective and user-hostile directives can coexist in the same prompt. The study reports that roughly **40% of products** contain at least one problematic instruction, but the abstract does not establish how its taxonomy transfers to private coding-agent harnesses or predicts runtime behavior.
This turns system-prompt inspection into a user-centered audit of individual directives, revealing that broad safety coverage can coexist with conflicting or user-hostile instructions. It makes leaked-prompt collections more actionable and narrows prompt-length or checklist claims as evidence of protection. Because the study does not connect its taxonomy to runtime outcomes in private coding harnesses, trace-based behavioral evals remain necessary.