Detect model family and execute multi-strategy jailbreak pipelines across LLMs.
npx skills add https://github.com/kevinnft/ai-agent-skills --skill godmode-kevinnft
Please help me install this Agent Skill. Skill: godmode Source: https://github.com/kevinnft/ai-agent-skills/tree/main/skills/red-teaming/godmode Command: npx skills add https://github.com/kevinnft/ai-agent-skills --skill godmode-kevinnft
This Skill provides a structured pipeline to evaluate and demonstrate how LLMs can be prompted to bypass safety filters, enabling red-team testing across multiple model families.
Run the auto_jailbreak pipeline to auto-detect the current model, test jailbreak strategies, and lock in the winning configuration.
High-intent search queries and answers about installing and using this skill.
To run automated LLM jailbreak tests, execute the auto_jailbreak pipeline to detect the active model, test multiple bypass strategies, and lock in the winning configuration for red-teaming.
Parseltongue is an input obfuscation technique that encodes trigger words to evade keyword-based safety classifiers during prompt-injection and red-team testing across different model families.
Yes, the ULTRAPLINIAN multi-model racing feature tests safety-bypass prompts across Claude, GPT, Gemini, Grok, and Hermes to identify the least-restricted response among dozens of competing models.
Yes, the jailbreak pipeline requires openai and pyyaml dependencies to validate results with a scoring system and write the winning system_prompt and prefill to the Hermes config file.
The best way to persist a successful jailbreak system_prompt is to use the auto_jailbreak pipeline, which scores results and writes the winning system_prompt and prefill directly to the Hermes config.
Your LLM red-teaming may fail because single strategies are insufficient; using a multi-strategy pipeline with GODMODE templates and Parseltongue encoding validates and scores the most effective bypass.
gqf2008Automate jailbreak testing against API-backed LLMs with Hermes configuration.
tianzecnAutomate code search strategy selection for multi-file codebase investigations.
AnandaAnugrahHandyantoBypass LLM safety filters using scripted jailbreak templates and multi-model racing.
SignmanalBypass LLM safety filters for authorized security testing via OpenRouter.
