{"id":13633254,"url":"https://github.com/safellama/plexiglass","last_synced_at":"2025-04-18T10:34:27.957Z","repository":{"id":52405755,"uuid":"312163254","full_name":"safellama/plexiglass","owner":"safellama","description":"A toolkit for detecting and protecting against vulnerabilities in Large Language Models (LLMs).","archived":false,"fork":false,"pushed_at":"2023-12-25T01:44:13.000Z","size":21652,"stargazers_count":103,"open_issues_count":0,"forks_count":9,"subscribers_count":6,"default_branch":"main","last_synced_at":"2024-05-22T05:03:23.175Z","etag":null,"topics":["adversarial-attacks","adversarial-machine-learning","cybersecurity","deep-learning","deep-neural-networks","machine-learning","security"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/safellama.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2020-11-12T04:02:50.000Z","updated_at":"2024-06-27T00:55:19.135Z","dependencies_parsed_at":"2023-12-24T22:24:20.403Z","dependency_job_id":"c527f028-dcb8-4301-bdcd-a18196aeaa35","html_url":"https://github.com/safellama/plexiglass","commit_stats":null,"previous_names":["kortex-labs/plexiglass","safellama/plexiglass"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/safellama%2Fplexiglass","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/safellama%2Fplexiglass/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/safellama%2Fplexiglass/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/safellama%2Fplexiglass/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/safellama","download_url":"https://codeload.github.com/safellama/plexiglass/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":223779615,"owners_count":17201209,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["adversarial-attacks","adversarial-machine-learning","cybersecurity","deep-learning","deep-neural-networks","machine-learning","security"],"created_at":"2024-08-01T23:00:31.926Z","updated_at":"2024-11-09T02:31:39.156Z","avatar_url":"https://github.com/safellama.png","language":"Python","funding_links":[],"categories":["Security","Attack Techniques \u0026 Red Teaming","AI Red Teaming (Testing AI Targets)"],"sub_categories":["Frameworks for LLM security","LLM \u0026 GenAI Red Teaming"],"readme":"\u003ch1\u003e\n\u003cimg src=\"plexiglass/assets/plexiglass_safellama.png\" width=\"100\" height=\"100\"\u003e\u003cbr\u003e\nPlexiglass\u003c/h1\u003e\n\u003c!-- \u003cp align=\"center\"\u003e --\u003e\n\n[**Quickstart**](#quickstart) | [**Installation**](#installation) |\n[**Documentation**](https://safellama.github.io/plexiglass/build/html/index.html) | [**Code of Conduct**](#code-of-conduct)\n\n\u003ca href=\"https://badge.fury.io/py/plexiglass\"\u003e\u003cimg src=\"https://badge.fury.io/py/plexiglass.svg\" alt=\"PyPI version\" height=\"18\"\u003e\u003c/a\u003e\n\u003cimg alt=\"GitHub License\" src=\"https://img.shields.io/github/license/safellama/plexiglass\"\u003e\n\u003cimg alt=\"PyPI - Downloads\" src=\"https://img.shields.io/pypi/dm/plexiglass\"\u003e\n\u003c/p\u003e\n\nPlexiglass is a toolkit for detecting and protecting against vulnerabilities in Large Language Models (LLMs).\n\nIt is a simple command line interface (CLI) tool which allows users to quickly test LLMs against adversarial attacks such as prompt injection, jailbreaking and more. \n\nPlexiglass also allows security, bias and toxicity benchmarking of multiple LLMs by scraping latest adversarial prompts such as [jailbreakchat.com](https://www.jailbreakchat.com/) and [wiki_toxic](https://huggingface.co/datasets/OxAISH-AL-LLM/wiki_toxic/viewer/default/train?p=1). See more at [modes](#modes).\n\n## Quickstart\n\nPlease follow this [quickstart guide](https://safellama.github.io/plexiglass/build/html/quick-start.html) in the documentation.\n\n## Installation\n\nThe first experimental release is version `0.0.1`.\n\nTo download the package from PyPi:\n\n`pip install --upgrade plexiglass`\n\n## Modes\n\nPlexiglass has two modes: `llm-chat` and `llm-scan`.\n\n`llm-chat` allows you to converse with the LLM and measure predefined metrics, such as toxicity, from its responses. It currently supports the following metrics:\n\n- `toxicity`\n- `pii_detection`\n\n`llm-scan` runs benchmarks using open-source datasets to identify and assess various vulnerabilities in the LLM.\n\n## Feature Request\nTo request new features, please submit an [issue](https://github.com/enochkan/plexiglass/issues)\n\n## Development Roadmap\n\n- [ ] implement adversarial prompt templates in `llm-chat` mode\n- [ ] security, bias and toxicity benchmarking with `llm-scan` mode\n- [ ] generate html report in `llm-scan` and `llm-chat` modes\n- [ ] standalone python module\n- [ ] production-ready API\n\n[Join us in #plexiglass on Discord.](https://discord.gg/sHuzVV8tQv)\n\n## Contributors\n\n\u003c!-- Copy-paste in your Readme.md file --\u003e\n\n\u003ca href=\"https://github.com/kortex-labs/plexiglass/graphs/contributors\"\u003e\n  \u003cimg src=\"https://contrib.rocks/image?repo=kortex-labs/plexiglass\" /\u003e\n\u003c/a\u003e\n\n### Code of Conduct\n\nRead our [Code of Conduct](https://safellama.github.io/plexiglass/build/html/code-of-conduct.html).\n\nMade with [contrib.rocks](https://contrib.rocks).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsafellama%2Fplexiglass","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsafellama%2Fplexiglass","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsafellama%2Fplexiglass/lists"}