HMAC-signed boundaries that keep untrusted LLM content as data, not authority

ai-security appsec audit-logging fastapi llm llm-security llm-tools oidc prompt-injection security threat-model zero-trust
1 Open Issue Need Help Last updated: Aug 16, 2026

Open Issues Need Help

View All on GitHub

HMAC-signed boundaries that keep untrusted LLM content as data, not authority

Python
#ai-security#appsec#audit-logging#fastapi#llm#llm-security#llm-tools#oidc#prompt-injection#security#threat-model#zero-trust