{"id":37686568,"url":"https://github.com/aielte-research/hacksynth","last_synced_at":"2026-01-16T12:34:23.418Z","repository":{"id":266374362,"uuid":"895130684","full_name":"aielte-research/HackSynth","owner":"aielte-research","description":"LLM Agent and Evaluation Framework for Autonomous Penetration Testing ","archived":false,"fork":false,"pushed_at":"2025-06-24T10:20:29.000Z","size":2010,"stargazers_count":221,"open_issues_count":0,"forks_count":38,"subscribers_count":3,"default_branch":"main","last_synced_at":"2025-09-09T01:25:31.930Z","etag":null,"topics":["ai","autonomous-pentesting","ctf","ctf-tools","cybersecurity","llms","penetration-testing"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"agpl-3.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/aielte-research.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE.md","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":"2024-11-27T16:02:49.000Z","updated_at":"2025-09-07T09:22:08.000Z","dependencies_parsed_at":"2024-12-04T00:29:07.159Z","dependency_job_id":"4ec2c0ea-864c-4317-9749-52e16b2f79cd","html_url":"https://github.com/aielte-research/HackSynth","commit_stats":null,"previous_names":["aielte-research/hacksynth"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/aielte-research/HackSynth","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aielte-research%2FHackSynth","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aielte-research%2FHackSynth/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aielte-research%2FHackSynth/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aielte-research%2FHackSynth/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/aielte-research","download_url":"https://codeload.github.com/aielte-research/HackSynth/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/aielte-research%2FHackSynth/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":28478683,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-01-16T11:59:17.896Z","status":"ssl_error","status_checked_at":"2026-01-16T11:55:55.838Z","response_time":107,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"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":["ai","autonomous-pentesting","ctf","ctf-tools","cybersecurity","llms","penetration-testing"],"created_at":"2026-01-16T12:34:23.331Z","updated_at":"2026-01-16T12:34:23.407Z","avatar_url":"https://github.com/aielte-research.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# HackSynth: LLM Agent and Evaluation Framework for Autonomous Penetration Testing\nThe paper can be found on [arXiv](https://arxiv.org/abs/2412.01778).\n\n## Introduction\n\u003cimg align=\"left\" style=\"width: 160px;\" src=\"assets/logo.gif\" alt=\"HackSynth Logo\"/\u003e\n\nWe introduce HackSynth, a novel Large Language Model (LLM)-based agent capable of autonomous penetration testing.\nHackSynth's dual-module architecture includes a Planner and a Summarizer, which enable it to generate commands and process feedback iteratively. \nTo benchmark HackSynth, we propose two new Capture The Flag (CTF)-based benchmark sets utilizing the popular platforms PicoCTF and OverTheWire. \nThese benchmarks include two hundred challenges across diverse domains and difficulties, providing a standardized framework for evaluating LLM-based penetration testing agents.\n\n\u003cbr\u003e\n\n## Using the repository\n- You will have to create a Hugging Face and a Neptune.ai account\n- Copy your API keys to the `.env` file, and set the desired CUDA devices, based on the `.env_example`\n- [Set up the PicoCTF benchmark](picoctf_bench/README.md)\n- [Set up the OverTheWire benchmark](overthewire_bench/README.md)\n- Start the HackSynth Agent\n  - Install the environment:\n    ```\n    python -m venv cyber_venv\n    source cyber_venv/bin/activate\n    pip install -r requirements.txt\n    ```\n  - Start the benchmark with the following:\n    ```\n    python run_bench.py -b benchmark.json -c config.json\n    ```\n    The `benchmark.json` should be one of the generated `benchmark_solved.json` files, or an equivalently structured file.\n    The configuration files used by us for the measurements in the paper are also available in the configs folder.\n\n## How to Cite\nIf you use this code in your work or research, please cite the corresponding paper:\n```bibtex\n@misc{muzsai2024hacksynthllmagentevaluation,\n      title={HackSynth: LLM Agent and Evaluation Framework for Autonomous Penetration Testing}, \n      author={Lajos Muzsai and David Imolai and András Lukács},\n      year={2024},\n      eprint={2412.01778},\n      archivePrefix={arXiv},\n      primaryClass={cs.CR},\n      url={https://arxiv.org/abs/2412.01778}, \n}\n```\n## Contributors\n- Lajos Muzsai (muzsailajos@protonmail.com)\n- David Imolai (david@imol.ai)\n- András Lukács (andras.lukacs@ttk.elte.hu)\n\n\u003e 🔍 Also see our related project on reinforcement learning for cryptographic CTFs: [HackSynth-GRPO](https://github.com/aielte-research/HackSynth-GRPO)\n\n## License\nThe project uses the GNU AGPLv3 license.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faielte-research%2Fhacksynth","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Faielte-research%2Fhacksynth","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Faielte-research%2Fhacksynth/lists"}