Apply AI/ML techniques to solve CTF challenges.
npx skills add https://github.com/a398275770-ctrl/ctf-skills --skill ctf-ai-ml-a398275770-ctrl
Please help me install this Agent Skill. Skill: ctf-ai-ml Source: https://github.com/a398275770-ctrl/ctf-skills/tree/main/ctf-ai-ml Command: npx skills add https://github.com/a398275770-ctrl/ctf-skills --skill ctf-ai-ml-a398275770-ctrl
This Skill provides a comprehensive set of AI and machine learning techniques specifically designed for solving CTF challenges, covering areas like adversarial examples, model inversion, and LLM attacks.
To analyze the model at 'model.pt', run the command: inspect_model 'model.pt'.
High-intent search queries and answers about installing and using this skill.
Adversarial example generation creates targeted perturbations to fool machine learning classifiers. This Skill automates crafting inputs that manipulate model outputs, specifically optimized for AI puzzles encountered in CTF competitions.
Model inversion reverse-engineers original inputs from model outputs using libraries like torch and numpy. This Skill performs inversion attacks to extract sensitive training data or reconstruct inputs from provided model files.
Yes, LLM attacks exploit vulnerabilities in language models, while membership inference determines if specific data was in the training set. This Skill uses transformers to execute both attack vectors against target models.
Yes, AI-ML CTF attacks require Python 3 with specific libraries including torch, transformers, numpy, scipy, and scikit-learn. The Skill also uses safetensors and Pillow for comprehensive model analysis and manipulation.
To analyze a model file, use the inspect_model command with your model path, such as 'inspect_model model.pt'. This initiates the Skill's automated analysis to identify vulnerabilities and attack surfaces.
Yes, the Skill is optimized for web exploitation, binary pwn, crypto, reverse engineering, forensics, and OSINT challenges. It integrates AI-ML attacks with these traditional CTF categories for comprehensive challenge solving.
DexterJieGenerate adversarial examples and attack ML models in CTF challenges.
zhnnky329Run robustness, sensitivity, error, and baseline checks on modeling outputs.
KongGithubDevCoordinates CTF skills across categories to analyze files and exploit services.
langbyyiGenerate adversarial examples and bypass LLM safety filters for CTF challenges.
