HN Showcase
Back to browse
GitHub Repository

Extended Threat Reasoning Architecture

0 starsPython

Xtra – a Python framework for reasoning about AI system threats

by reboy·Jun 26, 2026·1 point·0 comments
Visit ProjectView on HN

AI Analysis

●●SolidBig BrainNiche Gem

FSM-based social engineering detection without LLMs is a genuinely contrarian approach.

Strengths
  • •State transition modeling catches attack patterns that content-based ML misses.
  • •Tracks specific signals like flattery density and asymmetry spikes across turns.
  • •Zero dependencies means fast inference with no model hosting costs.
Weaknesses
  • •Zero stars and no test suite makes real-world reliability impossible to verify.
  • •Only handles conversational attacks, not broader AI threat categories.
Category
Security
Target Audience

Security engineers building conversational AI systems

Similar To

Garak · LLM Guard · PromptInject

Similar Projects

Security●●Solid

StegoForge – A zero-dependency steganography and forensics engine

Zero-dependency binaries with Twitter recompression survival testing built in.

WizardryNiche Gem
nour833
202mo ago
Security●●Solid

ClawGuard – Detect 42 prompt injection patterns in <10ms

42 prompt injection patterns detected in under 10ms with zero dependencies.

Niche GemShip It
joergmichno
233mo ago
Developer Tools●●Solid

SpecFact CLI – reverse engineer Python repos into enforceable specs

This repo actually wires the specfact CLI to a tiny, reproducible codebase so you can import-from-code, generate .specfact bundles, and run enforcement presets with one-liners. The backlog-sync adapter and a deliberately buggy sidecar demo make failure modes easy to exercise, and the README lists exact smoke commands and logs to verify results. Inferred specs will always risk false positives, but the project shows practical artifacts (change_tracking, results logs) rather than theory.

Niche GemSolve My Problem
noldai
114mo ago