{"data":{"slug":"safellama-plexiglass","name":"plexiglass","tagline":"A toolkit for detecting and protecting against vulnerabilities in Large Language Models (LLMs).","github_url":"https://github.com/w4-advisory/plexiglass","owner":"w4-advisory","repo":"plexiglass","owner_avatar_url":"https://avatars.githubusercontent.com/u/141878442?v=4","primary_language":"Python","stars":156,"forks":18,"topics":["adversarial-attacks","adversarial-machine-learning","cybersecurity","deep-learning","deep-neural-networks","machine-learning","security"],"archived":false,"github_pushed_at":"2026-02-04T22:18:58+00:00","maintenance_label":"Slowing","stars_delta_30d":3,"url":"https://www.graphcanon.com/tools/safellama-plexiglass","markdown_url":"https://www.graphcanon.com/tools/safellama-plexiglass.md","api_url":"https://www.graphcanon.com/api/graphcanon/tools/safellama-plexiglass","graph_url":"https://www.graphcanon.com/api/graphcanon/graph?tool=safellama-plexiglass","description":"A toolkit for detecting and protecting against vulnerabilities in Large Language Models (LLMs).","homepage_url":null,"license":"Apache-2.0","open_issues":0,"watchers":5,"ai_summary":"Plexiglass is a Python-based toolkit designed to detect and mitigate vulnerabilities in Large Language Models, focusing on adversarial attacks and cybersecurity.","readme_excerpt":"## Installation\n\nThe first experimental release is version `0.0.1`.\n\nTo download the package from PyPi:\n\n`pip install --upgrade plexiglass`","github_created_at":"2020-11-12T04:02:50+00:00","created_at":"2026-07-11T23:14:10.108165+00:00","updated_at":"2026-09-20T04:30:12.993646+00:00","categories":[{"slug":"evaluation-observability","name":"Evaluation & Observability","url":"https://www.graphcanon.com/categories/evaluation-observability","markdown_url":"https://www.graphcanon.com/categories/evaluation-observability.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/evaluation-observability"}],"tags":[{"slug":"adversarial-attacks","name":"adversarial-attacks"},{"slug":"adversarial-machine-learning","name":"adversarial-machine-learning"},{"slug":"cybersecurity","name":"cybersecurity"},{"slug":"deep-learning","name":"deep-learning"},{"slug":"deep-neural-networks","name":"deep-neural-networks"},{"slug":"machine-learning","name":"machine-learning"},{"slug":"security","name":"security"}],"trust":{"provenance":{"is_fork":false,"github_id":312163254,"owner_type":"Organization","methodology":"github_public_v1","parent_repo":null,"near_duplicate_slugs":[]},"computed_at":"2026-09-18T08:21:49.907Z","maintenance":{"label":"Slowing","score":36,"methodology":"github_public_v1","releases_90d":0,"days_since_push":225,"last_release_at":null,"stars_delta_30d":3,"open_issues_delta_30d":0},"security_summary":{"status":"no_lockfile","scanner":null,"low_count":0,"high_count":0,"last_scan_at":"2026-09-18T08:21:52.186Z","medium_count":0,"scan_profile":"none","critical_count":0}},"capability_facts":{"scan":{"source":"repo_scan","observed_at":"2026-09-18T08:21:48.858Z"},"has_cli":{"value":true,"source":"pyproject.toml:[project.scripts]","observed_at":"2026-09-18T08:21:48.858Z"},"languages":{"value":["python"],"source":"github.language+pyproject.toml","observed_at":"2026-09-18T08:21:48.858Z"},"license_spdx":{"value":"Apache-2.0","source":"github.license","observed_at":"2026-09-18T08:21:48.858Z"}},"decision_facts":{"hosting":null,"pricing":{"model":"freemium","summary":"The toolkit is free to use under the Apache-2.0 license, but advanced features or support might incur costs."},"requirements":{"notes":["Python environment is required for installation and operation."],"min_ram_gb":4,"requires_docker":false},"constraints":{"min_ram_gb":4,"pricing_model":"freemium","requires_docker":false},"when_to_use":["When you need to evaluate the robustness of Large Language Models against adversarial attacks.","If your project involves deep learning models and you require tools to enhance cybersecurity measures specifically for LLMs."],"when_not_to_use":["If your project does not involve Large Language Models and you are looking for general cybersecurity tools.","When you are working with small-scale models where the overhead of using Plexiglass might not be justified."],"source":"enrich:decision_facts","observed_at":"2026-09-18T11:07:20.568Z"},"constraint_facets":{"min_ram_gb":4,"pricing_model":"freemium","requires_docker":false},"decision_summary":[{"label":"Pricing","value":"freemium - The toolkit is free to use under the Apache-2.0 license, but advanced features or support might incur costs."},{"label":"Requirements","value":"Min 4 GB RAM; Python environment is required for installation and operation."},{"label":"Adopt for","value":"Plexiglass is a Python-based toolkit aimed at detecting and mitigating vulnerabilities in Large Language Models, with a focus on cybersecurity and adversarial attacks."},{"label":"License detail","value":"Plexiglass is available under the Apache-2.0 license, allowing for broad usage and modification with attribution."}]}}