Karl Pertsch

68 papers A* 7A 1Misc 11Journal 41Unranked 8
YearRankTypeTitle / Venue / Authors
2026 J jnl
CoRR
Marcel Torne, Karl Pertsch, Homer Walke, Kyle Vedder, Suraj Nair, Brian Ichter, Allen Z. Ren, Haohuan Wang, Jiaming Tang, Kyle Stachowicz, Karan Dhabalia, Michael Equi, Quan Vuong, Jost Tobias Springenberg, Sergey Levine, Chelsea Finn, Danny Driess
2026 J jnl
CoRR
Tony Lee, Andrew Wagenmaker, Karl Pertsch, Percy Liang, Sergey Levine, Chelsea Finn
2026 J jnl
CoRR
William Chen, Jagdeep Singh Bhatia, Catherine Glossop, Nikhil Mathihalli, Ria Doshi, Andy Tang, Danny Driess, Karl Pertsch, Sergey Levine
2025 A conf
IROS
Olivia Y. Lee, Annie Xie, Kuan Fang, Karl Pertsch, Chelsea Finn
2025 J jnl
CoRR
Zhiyuan Zhou, Pranav Atreya, You Liang Tan, Karl Pertsch, Sergey Levine
2025 J jnl
CoRR
Simar Kareer, Karl Pertsch, James Darpinian, Judy Hoffman, Danfei Xu, Sergey Levine, Chelsea Finn, Suraj Nair
2025 J jnl
CoRR
Karl Pertsch, Kyle Stachowicz, Brian Ichter, Danny Driess, Suraj Nair, Quan Vuong, Oier Mees, Chelsea Finn, Sergey Levine
2025 A* conf
ICML
Lucy Xiaoyang Shi, Brian Ichter, Michael Robert Equi, Liyiming Ke, Karl Pertsch, Quan Vuong, James Tanner, Anna Walling, Haohuan Wang, Niccolo Fusai, Adrian Li-Bell, Danny Driess, Lachy Groom, Sergey Levine, Chelsea Finn
2025 J jnl
CoRR
Lucy Xiaoyang Shi, Brian Ichter, Michael Equi, Liyiming Ke, Karl Pertsch, Quan Vuong, James Tanner, Anna Walling, Haohuan Wang, Niccolo Fusai, Adrian Li-Bell, Danny Driess, Lachy Groom, Sergey Levine, Chelsea Finn
2025 J jnl
CoRR
Danny Driess, Jost Tobias Springenberg, Brian Ichter, Lili Yu, Adrian Li-Bell, Karl Pertsch, Allen Z. Ren, Homer Walke, Quan Vuong, Lucy Xiaoyang Shi, Sergey Levine
2025 J jnl
CoRR
Arhan Jain, Mingtong Zhang, Kanav Arora, William Chen, Marcel Torne, Muhammad Zubair Irshad, Sergey Zakharov, Yue Wang, Sergey Levine, Chelsea Finn, Wei-Chiu Ma, Dhruv Shah, Abhishek Gupta, Karl Pertsch
2025 J jnl
CoRR
Pranav Atreya, Karl Pertsch, Tony Lee, Moo Jin Kim, Arhan Jain, Artur Kuramshin, Clemens Eppner, Cyrus Neary, Edward S. Hu, Fabio Ramos, Jonathan Tremblay, Kanav Arora, Kirsty Ellis, Luca Macesanu, Matthew Leonard, Meedeum Cho, Ozgur Aslan, Shivin Dass, Jie Wang, Xingfang Yuan, Xuning Yang, Abhishek Gupta, Dinesh Jayaraman, Glen Berseth, Kostas Daniilidis, Roberto Martin Martin, Youngwoon Lee, Percy Liang, Chelsea Finn, Sergey Levine
2025 J jnl
CoRR
Yajat Yadav, Zhiyuan Zhou, Andrew Wagenmaker, Karl Pertsch, Sergey Levine
2025 J jnl
CoRR
William Chen, Suneel Belkhale, Suvir Mirchandani, Oier Mees, Danny Driess, Karl Pertsch, Sergey Levine
2025 J jnl
CoRR
Physical Intelligence, Kevin Black, Noah Brown, James Darpinian, Karan Dhabalia, Danny Driess, Adnan Esmail, Michael Equi, Chelsea Finn, Niccolo Fusai, Manuel Y. Galliker, Dibya Ghosh, Lachy Groom, Karol Hausman, Brian Ichter, Szymon Jakubczak, Tim Jones, Liyiming Ke, Devin LeBlanc, Sergey Levine, Adrian Li-Bell, Mohith Mothukuri, Suraj Nair, Karl Pertsch, Allen Z. Ren, Lucy Xiaoyang Shi, Laura Smith, Jost Tobias Springenberg, Kyle Stachowicz, James Tanner, Quan Vuong, Homer Walke, Anna Walling, Haohuan Wang, Lili Yu, Ury Zhilinsky
2025 J jnl
CoRR
Physical Intelligence, Ali Amin, Raichelle Aniceto, Ashwin Balakrishna, Kevin Black, Ken Conley, Grace Connors, James Darpinian, Karan Dhabalia, Jared DiCarlo, Danny Driess, Michael Equi, Adnan Esmail, Yunhao Fang, Chelsea Finn, Catherine Glossop, Thomas Godden, Ivan Goryachev, Lachy Groom, Hunter Hancock, Karol Hausman, Gashon Hussein, Brian Ichter, Szymon Jakubczak, Rowan Jen, Tim Jones, Ben Katz, Liyiming Ke, Chandra Kuchi, Marinda Lamb, Devin LeBlanc, Sergey Levine, Adrian Li-Bell, Yao Lu, Vishnu Mano, Mohith Mothukuri, Suraj Nair, Karl Pertsch, Allen Z. Ren, Charvi Sharma, Lucy Xiaoyang Shi, Laura Smith, Jost Tobias Springenberg, Kyle Stachowicz, Will Stoeckle, Alex Swerdlow, James Tanner, Marcel Torne, Quan Vuong, Anna Walling, Haohuan Wang, Blake Williams, Sukwon Yoo, Lili Yu, Ury Zhilinsky, Zhiyuan Zhou
2024 J jnl
CoRR
Olivia Y. Lee, Annie Xie, Kuan Fang, Karl Pertsch, Chelsea Finn
2024 conf
Robotics: Science and Systems
Alexander Khazatsky, Karl Pertsch, Suraj Nair, Ashwin Balakrishna, Sudeep Dasari, Siddharth Karamcheti, Soroush Nasiriany, Mohan Kumar Srirama, Lawrence Yunliang Chen, Kirsty Ellis, Peter David Fagan, Joey Hejna, Masha Itkina, Marion Lepert, Yecheng Jason Ma, Patrick Tree Miller, Jimmy Wu, Suneel Belkhale, Shivin Dass, Huy Ha, Arhan Jain, Abraham Lee, Youngwoon Lee, Marius Memmel, Sungjae Park, Ilija Radosavovic, Kaiyuan Wang, Albert Zhan, Kevin Black, Cheng Chi, Kyle Beltran Hatch, Shan Lin, Jingpei Lu, Jean Mercat, Abdul Rehman, Pannag R. Sanketi, Archit Sharma, Cody Simpson, Quan Vuong, Homer Rich Walke, Blake Wulfe, Ted Xiao, Jonathan Heewon Yang, Arefeh Yavary, Tony Z. Zhao, Christopher Agia, Rohan Baijal, Mateo Guaman Castro, Daphne Chen, Qiuyu Chen, Trinity Chung, Jaimyn Drake, Ethan Paul Foster, Jensen Gao, David Antonio Herrera, Minho Heo, Kyle Hsu, Jiaheng Hu, Donovon Jackson, Charlotte Le, Yunshuang Li, Roy Lin, Zehan Ma, Abhiram Maddukuri, Suvir Mirchandani, Daniel Morton, Tony Nguyen, Abigail O'Neill, Rosario Scalise, Derick Seale, Victor Son, Stephen Tian, Emi Tran, Andrew E. Wang, Yilin Wu, Annie Xie, Jingyun Yang, Patrick Yin, Yunchu Zhang, Osbert Bastani, Glen Berseth, Jeannette Bohg, Ken Goldberg, Abhinav Gupta, Abhishek Gupta, Dinesh Jayaraman, Joseph J. Lim, Jitendra Malik, Roberto Martín-Martín, Subramanian Ramamoorthy, Dorsa Sadigh, Shuran Song, Jiajun Wu, Michael C. Yip, Yuke Zhu, Thomas Kollar, Sergey Levine, Chelsea Finn
2024 J jnl
CoRR
Alexander Khazatsky, Karl Pertsch, Suraj Nair, Ashwin Balakrishna, Sudeep Dasari, Siddharth Karamcheti, Soroush Nasiriany, Mohan Kumar Srirama, Lawrence Yunliang Chen, Kirsty Ellis, Peter David Fagan, Joey Hejna, Masha Itkina, Marion Lepert, Jason Ma, Patrick Tree Miller, Jimmy Wu, Suneel Belkhale, Shivin Dass, Huy Ha, Arhan Jain, Abraham Lee, Youngwoon Lee, Marius Memmel, Sungjae Park, Ilija Radosavovic, Kaiyuan Wang, Albert Zhan, Kevin Black, Cheng Chi, Kyle Beltran Hatch, Shan Lin, Jingpei Lu, Jean Mercat, Abdul Rehman, Pannag R. Sanketi, Archit Sharma, Cody Simpson, Quan Vuong, Homer Rich Walke, Blake Wulfe, Ted Xiao, Jonathan Heewon Yang, Arefeh Yavary, Tony Z. Zhao, Christopher Agia, Rohan Baijal, Mateo Guaman Castro, Daphne Chen, Qiuyu Chen, Trinity Chung, Jaimyn Drake, Ethan Paul Foster, et al.
2024 Misc conf
CoRL
Xuanlin Li, Kyle Hsu, Jiayuan Gu, Oier Mees, Karl Pertsch, Homer Rich Walke, Chuyuan Fu, Ishikaa Lunawat, Isabel Sieh, Sean Kirmani, Sergey Levine, Jiajun Wu, Chelsea Finn, Hao Su, Quan Vuong, Ted Xiao
2024 J jnl
CoRR
Xuanlin Li, Kyle Hsu, Jiayuan Gu, Karl Pertsch, Oier Mees, Homer Rich Walke, Chuyuan Fu, Ishikaa Lunawat, Isabel Sieh, Sean Kirmani, Sergey Levine, Jiajun Wu, Chelsea Finn, Hao Su, Quan Vuong, Ted Xiao
2024 conf
Robotics: Science and Systems
Dibya Ghosh, Homer Rich Walke, Karl Pertsch, Kevin Black, Oier Mees, Sudeep Dasari, Joey Hejna, Tobias Kreiman, Charles Xu, Jianlan Luo, You Liang Tan, Lawrence Yunliang Chen, Quan Vuong, Ted Xiao, Pannag R. Sanketi, Dorsa Sadigh, Chelsea Finn, Sergey Levine
2024 J jnl
CoRR
Octo Model Team, Dibya Ghosh, Homer Walke, Karl Pertsch, Kevin Black, Oier Mees, Sudeep Dasari, Joey Hejna, Tobias Kreiman, Charles Xu, Jianlan Luo, You Liang Tan, Lawrence Yunliang Chen, Pannag Sanketi, Quan Vuong, Ted Xiao, Dorsa Sadigh, Chelsea Finn, Sergey Levine
2024 A* conf
ICRA
Abby O'Neill, Abdul Rehman, Abhiram Maddukuri, Abhishek Gupta, Abhishek Padalkar, Abraham Lee, Acorn Pooley, Agrim Gupta, Ajay Mandlekar, Ajinkya Jain, Albert Tung, Alex Bewley, Alexander Herzog, Alex Irpan, Alexander Khazatsky, Anant Rai, Anchit Gupta, Andrew E. Wang, Anikait Singh, Animesh Garg, Aniruddha Kembhavi, Annie Xie, Anthony Brohan, Antonin Raffin, Archit Sharma, Arefeh Yavary, Arhan Jain, Ashwin Balakrishna, Ayzaan Wahid, Ben Burgess-Limerick, Beomjoon Kim, Bernhard Schölkopf, Blake Wulfe, Brian Ichter, Cewu Lu, Charles Xu, Charlotte Le, Chelsea Finn, Chen Wang, Chenfeng Xu, Cheng Chi, Chenguang Huang, Christine Chan, Christopher Agia, Chuer Pan, Chuyuan Fu, Coline Devin, Danfei Xu, Daniel Morton, Danny Driess, Daphne Chen, Deepak Pathak, Dhruv Shah, Dieter Büchler, Dinesh Jayaraman, Dmitry Kalashnikov, Dorsa Sadigh, Edward Johns, Ethan Paul Foster, Fangchen Liu, Federico Ceola, Fei Xia, Feiyu Zhao, Freek Stulp, Gaoyue Zhou, Gaurav S. Sukhatme, Gautam Salhotra, Ge Yan, Gilbert Feng, Giulio Schiavi, Glen Berseth, Gregory Kahn, Guanzhi Wang, Hao Su, Haoshu Fang, Haochen Shi, Henghui Bao, Heni Ben Amor, Henrik I. Christensen, Hiroki Furuta, Homer Walke, Hongjie Fang, Huy Ha, Igor Mordatch, Ilija Radosavovic, Isabel Leal, Jacky Liang, Jad Abou-Chakra, Jaehyung Kim, Jaimyn Drake, Jan Peters, Jan Schneider, Jasmine Hsu, Jeannette Bohg, Jeffrey T. Bingham, Jeffrey Wu, Jensen Gao, Jiaheng Hu, Jiajun Wu, Jialin Wu, Jiankai Sun, Jianlan Luo, Jiayuan Gu, Jie Tan, Jihoon Oh, Jimmy Wu, Jingpei Lu, Jingyun Yang, Jitendra Malik, João Silvério, Joey Hejna, Jonathan Booher, Jonathan Tompson, Jonathan Yang, Jordi Salvador, Joseph J. Lim, Junhyek Han, Kaiyuan Wang, Kanishka Rao, Karl Pertsch, Karol Hausman, Keegan Go, Keerthana Gopalakrishnan, Ken Goldberg, Kendra Byrne, Kenneth Oslund, Kento Kawaharazuka, Kevin Black, Kevin Lin, Kevin Zhang, Kiana Ehsani, Kiran Lekkala, Kirsty Ellis, Krishan Rana, Krishnan Srinivasan, Kuan Fang, Kunal Pratap Singh, Kuo-Hao Zeng, Kyle Hatch, Kyle Hsu, Laurent Itti, Lawrence Yunliang Chen, Lerrel Pinto, Li Fei-Fei, Liam Tan, Linxi Jim Fan, Lionel Ott, Lisa Lee, Luca Weihs, Magnum Chen, Marion Lepert, Marius Memmel, Masayoshi Tomizuka, Masha Itkina, Mateo Guaman Castro, Max Spero, Maximilian Du, Michael Ahn, Michael C. Yip, Mingtong Zhang, Mingyu Ding, Minho Heo, Mohan Kumar Srirama, Mohit Sharma, Moo Jin Kim, Naoaki Kanazawa, Nicklas Hansen, Nicolas Heess, Nikhil J. Joshi, Niko Sünderhauf, Ning Liu, Norman Di Palo, Nur Muhammad (Mahi) Shafiullah, Oier Mees, Oliver Kroemer, Osbert Bastani, Pannag R. Sanketi, Patrick Tree Miller, Patrick Yin, Paul Wohlhart, Peng Xu, Peter David Fagan, Peter Mitrano, Pierre Sermanet, Pieter Abbeel, Priya Sundaresan, Qiuyu Chen, Quan Vuong, Rafael Rafailov, Ran Tian, Ria Doshi, Roberto Martín-Martín, Rohan Baijal, Rosario Scalise, Rose Hendrix, Roy Lin, Runjia Qian, Ruohan Zhang, Russell Mendonca, Rutav Shah, Ryan Hoque, Ryan Julian, Samuel Bustamante-Gomez, Sean Kirmani, Sergey Levine, Shan Lin, Sherry Moore, Shikhar Bahl, Shivin Dass, Shubham D. Sonawani, Shuran Song, Sichun Xu, Siddhant Haldar, Siddharth Karamcheti, Simeon Adebola, Simon Guist, Soroush Nasiriany, Stefan Schaal, Stefan Welker, Stephen Tian, Subramanian Ramamoorthy, Sudeep Dasari, Suneel Belkhale, Sungjae Park, Suraj Nair, Suvir Mirchandani, Takayuki Osa, Tanmay Gupta, Tatsuya Harada, Tatsuya Matsushima, Ted Xiao, Thomas Kollar, Tianhe Yu, Tianli Ding, Todor Davchev, Tony Z. Zhao, Travis Armstrong, Trevor Darrell, Trinity Chung, Vidhi Jain, Vincent Vanhoucke, Wei Zhan, Wenxuan Zhou, Wolfram Burgard, Xi Chen, Xiaolong Wang, Xinghao Zhu, Xinyang Geng, Xiyuan Liu, Liangwei Xu, Xuanlin Li, Yao Lu, Yecheng Jason Ma, Yejin Kim, Yevgen Chebotar, Yifan Zhou, Yifeng Zhu, Yilin Wu, Ying Xu, Yixuan Wang, Yonatan Bisk, Yoonyoung Cho, Youngwoon Lee, Yuchen Cui, Yue Cao, Yueh-Hua Wu, Yujin Tang, Yuke Zhu, Yunchu Zhang, Yunfan Jiang, Yunshuang Li, Yunzhu Li, Yusuke Iwasawa, Yutaka Matsuo, Zehan Ma, Zhuo Xu, Zichen Jeff Cui, Zichen Zhang, Zipeng Lin
2024 Misc conf
CoRL
Moo Jin Kim, Karl Pertsch, Siddharth Karamcheti, Ted Xiao, Ashwin Balakrishna, Suraj Nair, Rafael Rafailov, Ethan Paul Foster, Pannag R. Sanketi, Quan Vuong, Thomas Kollar, Benjamin Burchfiel, Russ Tedrake, Dorsa Sadigh, Sergey Levine, Percy Liang, Chelsea Finn
2024 J jnl
CoRR
Moo Jin Kim, Karl Pertsch, Siddharth Karamcheti, Ted Xiao, Ashwin Balakrishna, Suraj Nair, Rafael Rafailov, Ethan Paul Foster, Grace Lam, Pannag Sanketi, Quan Vuong, Thomas Kollar, Benjamin Burchfiel, Russ Tedrake, Dorsa Sadigh, Sergey Levine, Percy Liang, Chelsea Finn
2024 J jnl
CoRR
Joey Hejna, Chethan Bhateja, Yichen Jiang, Karl Pertsch, Dorsa Sadigh
2024 Misc conf
CoRL
Joey Hejna, Chethan Anand Bhateja, Yichen Jiang, Karl Pertsch, Dorsa Sadigh
2024 Misc conf
CoRL
Michal Zawalski, William Chen, Karl Pertsch, Oier Mees, Chelsea Finn, Sergey Levine
2024 J jnl
CoRR
Michal Zawalski, William Chen, Karl Pertsch, Oier Mees, Chelsea Finn, Sergey Levine
2024 A* conf
ICRA
Jesse Zhang, Karl Pertsch, Jiahui Zhang, Joseph J. Lim
2024 conf
Robotics: Science and Systems
Lucy Xiaoyang Shi, Zheyuan Hu, Tony Z. Zhao, Archit Sharma, Karl Pertsch, Jianlan Luo, Sergey Levine, Chelsea Finn
2024 J jnl
CoRR
Lucy Xiaoyang Shi, Zheyuan Hu, Tony Z. Zhao, Archit Sharma, Karl Pertsch, Jianlan Luo, Sergey Levine, Chelsea Finn
2024 J jnl
CoRR
Kevin Black, Noah Brown, Danny Driess, Adnan Esmail, Michael Equi, Chelsea Finn, Niccolo Fusai, Lachy Groom, Karol Hausman, Brian Ichter, Szymon Jakubczak, Tim Jones, Liyiming Ke, Sergey Levine, Adrian Li-Bell, Mohith Mothukuri, Suraj Nair, Karl Pertsch, Lucy Xiaoyang Shi, James Tanner, Quan Vuong, Anna Walling, Haohuan Wang, Ury Zhilinsky
2023 Misc conf
CoRL
Jesse Zhang, Jiahui Zhang, Karl Pertsch, Ziyi Liu, Xiang Ren, Minsuk Chang, Shao-Hua Sun, Joseph J. Lim
2023 J jnl
CoRR
Jesse Zhang, Jiahui Zhang, Karl Pertsch, Ziyi Liu, Xiang Ren, Minsuk Chang, Shao-Hua Sun, Joseph J. Lim
2023 J jnl
CoRR
Taewook Nam, Juyong Lee, Jesse Zhang, Sung Ju Hwang, Joseph J. Lim, Karl Pertsch
2023 conf
Robotics: Science and Systems
Shivin Dass, Karl Pertsch, Hejia Zhang, Youngwoon Lee, Joseph J. Lim, Stefanos Nikolaidis
2023 Misc conf
CoRL
Yevgen Chebotar, Quan Vuong, Karol Hausman, Fei Xia, Yao Lu, Alex Irpan, Aviral Kumar, Tianhe Yu, Alexander Herzog, Karl Pertsch, Keerthana Gopalakrishnan, Julian Ibarz, Ofir Nachum, Sumedh Anand Sontakke, Grecia Salazar, Huong T. Tran, Jodilyn Peralta, Clayton Tan, Deeksha Manjunath, Jaspiar Singh, Brianna Zitkovich, Tomas Jackson, Kanishka Rao, Chelsea Finn, Sergey Levine
2023 J jnl
CoRR
Yevgen Chebotar, Quan Vuong, Alex Irpan, Karol Hausman, Fei Xia, Yao Lu, Aviral Kumar, Tianhe Yu, Alexander Herzog, Karl Pertsch, Keerthana Gopalakrishnan, Julian Ibarz, Ofir Nachum, Sumedh Sontakke, Grecia Salazar, Huong T. Tran, Jodilyn Peralta, Clayton Tan, Deeksha Manjunath, Jaspiar Singh, Brianna Zitkovich, Tomas Jackson, Kanishka Rao, Chelsea Finn, Sergey Levine
2023 conf
Robotics: Science and Systems
Anthony Brohan, Noah Brown, Justice Carbajal, Yevgen Chebotar, Joseph Dabis, Chelsea Finn, Keerthana Gopalakrishnan, Karol Hausman, Alexander Herzog, Jasmine Hsu, Julian Ibarz, Brian Ichter, Alex Irpan, Tomas Jackson, Sally Jesmonth, Nikhil J. Joshi, Ryan Julian, Dmitry Kalashnikov, Yuheng Kuang, Isabel Leal, Kuang-Huei Lee, Sergey Levine, Yao Lu, Utsav Malla, Deeksha Manjunath, Igor Mordatch, Ofir Nachum, Carolina Parada, Jodilyn Peralta, Emily Perez, Karl Pertsch, Jornell Quiambao, Kanishka Rao, Michael S. Ryoo, Grecia Salazar, Pannag R. Sanketi, Kevin Sayed, Jaspiar Singh, Sumedh Sontakke, Austin Stone, Clayton Tan, Huong T. Tran, Vincent Vanhoucke, Steve Vega, Quan Vuong, Fei Xia, Ted Xiao, Peng Xu, Sichun Xu, Tianhe Yu, Brianna Zitkovich
2023 Misc conf
CoRL
Brianna Zitkovich, Tianhe Yu, Sichun Xu, Peng Xu, Ted Xiao, Fei Xia, Jialin Wu, Paul Wohlhart, Stefan Welker, Ayzaan Wahid, Quan Vuong, Vincent Vanhoucke, Huong T. Tran, Radu Soricut, Anikait Singh, Jaspiar Singh, Pierre Sermanet, Pannag R. Sanketi, Grecia Salazar, Michael S. Ryoo, Krista Reymann, Kanishka Rao, Karl Pertsch, Igor Mordatch, Henryk Michalewski, Yao Lu, Sergey Levine, Lisa Lee, Tsang-Wei Edward Lee, Isabel Leal, Yuheng Kuang, Dmitry Kalashnikov, Ryan Julian, Nikhil J. Joshi, Alex Irpan, Brian Ichter, Jasmine Hsu, Alexander Herzog, Karol Hausman, Keerthana Gopalakrishnan, Chuyuan Fu, Pete Florence, Chelsea Finn, Kumar Avinava Dubey, Danny Driess, Tianli Ding, Krzysztof Marcin Choromanski, Xi Chen, Yevgen Chebotar, Justice Carbajal, Noah Brown, Anthony Brohan, Montserrat Gonzalez Arenas, Kehang Han
2023 J jnl
CoRR
Anthony Brohan, Noah Brown, Justice Carbajal, Yevgen Chebotar, Xi Chen, Krzysztof Choromanski, Tianli Ding, Danny Driess, Avinava Dubey, Chelsea Finn, Pete Florence, Chuyuan Fu, Montse Gonzalez Arenas, Keerthana Gopalakrishnan, Kehang Han, Karol Hausman, Alexander Herzog, Jasmine Hsu, Brian Ichter, Alex Irpan, Nikhil J. Joshi, Ryan Julian, Dmitry Kalashnikov, Yuheng Kuang, Isabel Leal, Lisa Lee, Tsang-Wei Edward Lee, Sergey Levine, Yao Lu, Henryk Michalewski, Igor Mordatch, Karl Pertsch, Kanishka Rao, Krista Reymann, Michael S. Ryoo, Grecia Salazar, Pannag Sanketi, Pierre Sermanet, Jaspiar Singh, Anikait Singh, Radu Soricut, Huong T. Tran, Vincent Vanhoucke, Quan Vuong, Ayzaan Wahid, Stefan Welker, Paul Wohlhart, Jialin Wu, Fei Xia, Ted Xiao, Peng Xu, Sichun Xu, Tianhe Yu, Brianna Zitkovich
2023 A* conf
NeurIPS
Sumedh Sontakke, Jesse Zhang, Sébastien M. R. Arnold, Karl Pertsch, Erdem Biyik, Dorsa Sadigh, Chelsea Finn, Laurent Itti
2023 J jnl
CoRR
Sumedh A. Sontakke, Jesse Zhang, Sébastien M. R. Arnold, Karl Pertsch, Erdem Biyik, Dorsa Sadigh, Chelsea Finn, Laurent Itti
2023 J jnl
CoRR
Jesse Zhang, Karl Pertsch, Jiahui Zhang, Joseph J. Lim
2022 Misc conf
CoRL
Karl Pertsch, Ruta Desai, Vikash Kumar, Franziska Meier, Joseph J. Lim, Dhruv Batra, Akshara Rai
2022 J jnl
CoRR
Karl Pertsch, Ruta Desai, Vikash Kumar, Franziska Meier, Joseph J. Lim, Dhruv Batra, Akshara Rai
2022 J jnl
CoRR
Shivin Dass, Karl Pertsch, Hejia Zhang, Youngwoon Lee, Joseph J. Lim, Stefanos Nikolaidis
2022 J jnl
CoRR
Anthony Brohan, Noah Brown, Justice Carbajal, Yevgen Chebotar, Joseph Dabis, Chelsea Finn, Keerthana Gopalakrishnan, Karol Hausman, Alexander Herzog, Jasmine Hsu, Julian Ibarz, Brian Ichter, Alex Irpan, Tomas Jackson, Sally Jesmonth, Nikhil J. Joshi, Ryan Julian, Dmitry Kalashnikov, Yuheng Kuang, Isabel Leal, Kuang-Huei Lee, Sergey Levine, Yao Lu, Utsav Malla, Deeksha Manjunath, Igor Mordatch, Ofir Nachum, Carolina Parada, Jodilyn Peralta, Emily Perez, Karl Pertsch, Jornell Quiambao, Kanishka Rao, Michael S. Ryoo, Grecia Salazar, Pannag Sanketi, Kevin Sayed, Jaspiar Singh, Sumedh Sontakke, Austin Stone, Clayton Tan, Huong T. Tran, Vincent Vanhoucke, Steve Vega, Quan Vuong, Fei Xia, Ted Xiao, Peng Xu, Sichun Xu, Tianhe Yu, Brianna Zitkovich
2022 A* conf
ICLR
Taewook Nam, Shao-Hua Sun, Karl Pertsch, Sung Ju Hwang, Joseph J. Lim
2022 J jnl
CoRR
Taewook Nam, Shao-Hua Sun, Karl Pertsch, Sung Ju Hwang, Joseph J. Lim
2022 A* conf
ICLR
Jun Yamada, Karl Pertsch, Anisha Gunjal, Joseph J. Lim
2022 J jnl
CoRR
Jun Yamada, Karl Pertsch, Anisha Gunjal, Joseph J. Lim
2021 Misc conf
CoRL
Karl Pertsch, Youngwoon Lee, Yue Wu, Joseph J. Lim
2021 J jnl
CoRR
Karl Pertsch, Youngwoon Lee, Yue Wu, Joseph J. Lim
2020 Misc conf
CoRL
Karl Pertsch, Youngwoon Lee, Joseph J. Lim
2020 J jnl
CoRR
Karl Pertsch, Youngwoon Lee, Joseph J. Lim
2020 conf
L4DC
Karl Pertsch, Oleh Rybkin, Jingyun Yang, Shenghao Zhou, Konstantinos G. Derpanis, Kostas Daniilidis, Joseph J. Lim, Andrew Jaegle
2020 A* conf
NeurIPS
Karl Pertsch, Oleh Rybkin, Frederik Ebert, Shenghao Zhou, Dinesh Jayaraman, Chelsea Finn, Sergey Levine
2020 J jnl
CoRR
Karl Pertsch, Oleh Rybkin, Frederik Ebert, Chelsea Finn, Dinesh Jayaraman, Sergey Levine
2020 Misc conf
CoRL
Jun Yamada, Youngwoon Lee, Gautam Salhotra, Karl Pertsch, Max Pflueger, Gaurav S. Sukhatme, Joseph J. Lim, Peter Englert
2020 J jnl
CoRR
Jun Yamada, Youngwoon Lee, Gautam Salhotra, Karl Pertsch, Max Pflueger, Gaurav S. Sukhatme, Joseph J. Lim, Peter Englert
2019 J jnl
CoRR
Karl Pertsch, Oleh Rybkin, Jingyun Yang, Konstantinos G. Derpanis, Joseph J. Lim, Kostas Daniilidis, Andrew Jaegle
2019 conf
ICLR (Poster)
Oleh Rybkin, Karl Pertsch, Konstantinos G. Derpanis, Kostas Daniilidis, Andrew Jaegle
2018 J jnl
CoRR
Oleh Rybkin, Karl Pertsch, Andrew Jaegle, Konstantinos G. Derpanis, Kostas Daniilidis
2018 conf
ACCV (3)
Omid Hosseini Jafari, Siva Karthik Mustikovela, Karl Pertsch, Eric Brachmann, Carsten Rother
2017 J jnl
CoRR
Omid Hosseini Jafari, Siva Karthik Mustikovela, Karl Pertsch, Eric Brachmann, Carsten Rother
APK_FEATURES_PDD-Tech_Annex.md
← Index APK_FEATURES_PDD-Tech_Annex.md markdown
# APK Extractor — Technical Annex

**Companion to:** APK_FEATURES_PDD.md
**Audience:** Engineers implementing the APK extractors
**Date:** 2026-02-21

---

## Table of Contents

1. [File Structure](#1-file-structure)
2. [Dependencies Setup](#2-dependencies-setup)
3. [Base Class: APKExtractor](#3-base-class-apkextractor)
4. [Dataclass Definitions](#4-dataclass-definitions)
5. [Tag Enum Updates](#5-tag-enum-updates)
6. [Extractor Implementation Details](#6-extractor-implementation-details)
7. [Ingestor Integration](#7-ingestor-integration)
8. [Hashes Dataclass Update](#8-hashes-dataclass-update)
9. [ClickHouse Schema](#9-clickhouse-schema)
10. [Androguard API Reference](#10-androguard-api-reference)
11. [Known Packer/Protector Signatures](#11-known-packerprotector-signatures)
12. [Sensitive API Categorization Reference](#12-sensitive-api-categorization-reference)

---

## 1. File Structure

Create the following files:

```
redb/extractors/
├── apk_extractor.py                          # APKExtractor base class
├── apk_extractors/
│   ├── __init__.py
│   ├── apk_features.py                       # APKFeaturesExtractor
│   ├── apk_manifest.py                       # APKManifestExtractor
│   ├── apk_permissions.py                    # APKPermissionsExtractor
│   ├── apk_signature.py                      # APKSignatureExtractor
│   ├── apk_dex.py                            # APKDexExtractor
│   ├── apk_resources.py                      # APKResourceExtractor
│   ├── apk_native_libs.py                    # APKNativeLibExtractor
│   └── apk_inconsistency_tests.py            # APKInconsistencyTestsExtractor
```

Modified files:

```
redb/extractors/enum.py                       # Add APK tags
redb/models/dataclasses.py                    # Add APK dataclasses + permhash field in Hashes
redb/ingestor.py                              # Wire up APK dispatch
requirements.txt                              # Add androguard, permhash
```

---

## 2. Dependencies Setup

Add to `requirements.txt`:

```
androguard>=4.1
permhash
```

Install and verify:

```bash
pip install androguard permhash
python -c "from androguard.core.apk import APK; print('androguard OK')"
python -c "import permhash; print('permhash OK')"
```

---

## 3. Base Class: APKExtractor

**File:** `redb/extractors/apk_extractor.py`

Follow the exact pattern of `PEExtractor` (`redb/extractors/pe_extractor.py`) and `MachOExtractor` (`redb/extractors/macho_extractor.py`).

```python
import logging
from abc import ABCMeta, abstractmethod
import zipfile

from androguard.core.apk import APK

from redb.extractors.extractor import Extractor

logger = logging.getLogger(__name__)


@abstractmethod
class APKExtractor(Extractor, metaclass=ABCMeta):

    def __init__(
        self,
        filepath,
        log,
        exporters=None,
        index_prefix=None,
        elastic_index=None,
        known_benign=False,
        known_malicious=False,
        apk=None,
    ):
        super().__init__(
            filepath,
            log,
            exporters,
            index_prefix,
            elastic_index,
            known_benign,
            known_malicious
        )
        self.apk = apk if apk else self._generate_apk_object()

    def _generate_apk_object(self):
        """Parse APK using androguard."""
        apk = None
        try:
            apk = APK(self.filepath)
            if not apk.is_valid_APK():
                self.log.warning(
                    f"APK validation warning for {self.hash.sha256}"
                )
                # Still return the object — partial parsing may still work
        except Exception as e:
            self.log.error(
                f"Format error parsing APK {self.hash.sha256}: {e}"
            )
        return apk

    def _is_valid_apk(self):
        """Check if the APK object was parsed successfully."""
        return self.apk is not None

    def _get_zip_file(self):
        """Get a zipfile.ZipFile handle for direct archive inspection."""
        try:
            return zipfile.ZipFile(self.filepath, 'r')
        except (zipfile.BadZipFile, Exception) as e:
            self.log.error(f"Failed to open APK as ZIP: {e}")
            return None

    def _list_files(self):
        """List all files in the APK archive."""
        if not self._is_valid_apk():
            return []
        try:
            return self.apk.get_files()
        except Exception as e:
            self.log.error(f"Error listing APK files: {e}")
            return []
```

**Key design decisions:**
- The `apk=` parameter allows the ingestor to parse once and share across all APK extractors (same pattern as `pe=` in PEExtractor, `macho=` in MachOExtractor, `elf=` in ELFExtractor)
- `_get_zip_file()` provides direct ZIP access for extractors that need it (resources, native libs, inconsistency tests)
- The androguard `APK` object is the primary parsing interface; `zipfile` is secondary for archive-level inspection

---

## 4. Dataclass Definitions

**File:** `redb/models/dataclasses.py`

Add after the Mach-O dataclasses section. Follow the exact style of existing dataclasses (type annotations, `Optional` for nullable fields, `field(default_factory=list)` for list defaults).

```python
# ============================================================================
# APK Dataclasses
# ============================================================================

@dataclass
class APKFeatures:
    """Core APK metadata and properties."""
    package_name: str
    app_name: str
    version_code: int
    version_name: str
    min_sdk_version: Optional[int] = None
    target_sdk_version: Optional[int] = None
    compile_sdk_version: Optional[int] = None
    main_activity: Optional[str] = None
    is_debuggable: bool = False
    allow_backup: bool = True
    uses_cleartext_traffic: bool = False
    supported_abis: List[str] = field(default_factory=list)
    dex_count: int = 0
    total_dex_size: int = 0
    total_file_count: int = 0
    has_native_code: bool = False
    has_assets: bool = False
    uses_libraries: List[str] = field(default_factory=list)
    earliest_content_modification: Optional[str] = None  # ISO 8601
    latest_content_modification: Optional[str] = None    # ISO 8601
    contains_embedded_apk: bool = False


@dataclass
class APKManifestComponent:
    """Single Android component (activity, service, receiver, provider)."""
    component_type: str      # "activity", "service", "receiver", "provider"
    class_name: str
    is_exported: bool
    intent_actions: List[str] = field(default_factory=list)
    intent_categories: List[str] = field(default_factory=list)


@dataclass
class APKManifest:
    """Full AndroidManifest.xml analysis."""
    activity_count: int
    service_count: int
    receiver_count: int
    provider_count: int
    activities: List[str] = field(default_factory=list)
    services: List[str] = field(default_factory=list)
    receivers: List[str] = field(default_factory=list)
    providers: List[str] = field(default_factory=list)
    exported_components: List[str] = field(default_factory=list)
    intent_filters_by_action: List[str] = field(default_factory=list)
    intent_filters_by_category: List[str] = field(default_factory=list)
    uses_features: List[str] = field(default_factory=list)
    meta_data: Optional[Dict] = None
    manifest_xml: Optional[str] = None  # Full decompiled XML text


@dataclass
class APKPermission:
    """Single permission entry for ClickHouse export."""
    permission_name: str
    protection_level: str    # "normal", "dangerous", "signature", "signatureOrSystem", "unknown"
    is_custom: bool = False  # True if declared by the app itself


@dataclass
class APKPermissions:
    """APK permission summary."""
    total_permission_count: int
    dangerous_permission_count: int
    permissions: List[str] = field(default_factory=list)
    dangerous_permissions: List[str] = field(default_factory=list)
    custom_permissions: List[str] = field(default_factory=list)
    permission_details: List[APKPermission] = field(default_factory=list)
    permhash: Optional[str] = None


@dataclass
class APKCertificate:
    """X.509 certificate from APK signature."""
    subject: str
    issuer: str
    serial_number: str
    valid_from: str
    valid_to: str
    thumbprint_sha1: str
    thumbprint_sha256: str
    algorithm: Optional[str] = None
    key_size: Optional[int] = None
    is_self_signed: bool = False


@dataclass
class APKCodeSigningInfo:
    """Complete APK signing information."""
    _id: str  # sha256
    is_signed: bool
    signature_scheme_versions: List[int] = field(default_factory=list)  # [1], [1,2], [1,2,3], etc.
    number_of_certificates: int = 0
    x509_certificates: Optional[List[APKCertificate]] = None
    signer_subject: Optional[str] = None
    signer_issuer: Optional[str] = None


@dataclass
class APKDexFile:
    """Analysis of a single DEX file."""
    filename: str
    sha256: str
    class_count: int
    method_count: int
    string_count: int
    top_packages: Optional[List[Dict]] = None  # [{"package": str, "class_count": int}]
    api_usage: Optional[Dict[str, List[str]]] = None  # {"reflection": [...], "crypto": [...], ...}
    obfuscation_indicators: Optional[Dict] = None


@dataclass
class APKResource:
    """Embedded resource file entry."""
    path: str
    size: int
    sha256: str
    filetype_magika: str
    is_suspicious: bool = False  # True for executable types


@dataclass
class APKNativeLib:
    """Native .so library entry."""
    abi: str          # e.g., "arm64-v8a"
    filename: str     # e.g., "libfoo.so"
    size: int
    sha256: str
    is_known_packer: bool = False  # True if matches known packer lib name


@dataclass
class APKInconsistencyTests:
    """APK anomaly and anti-analysis detection results."""
    test_zip_bomb: Optional[bool] = None
    test_zip_duplicate_entries: Optional[bool] = None
    test_zip_path_traversal: Optional[bool] = None
    test_zip_suspicious_timestamps: Optional[bool] = None
    test_hidden_dex_files: Optional[bool] = None
    test_manifest_component_mismatch: Optional[bool] = None
    test_debuggable_release: Optional[bool] = None
    test_emulator_detection_strings: Optional[bool] = None
    test_debugger_detection: Optional[bool] = None
    test_root_detection: Optional[bool] = None
```

---

## 5. Tag Enum Updates

**File:** `redb/extractors/enum.py`

Add at the end of the `Tag` class, after the ELF section:

```python
    # APK
    APK_FEATURES = "apk_features"
    APK_MANIFEST = "apk_manifest"
    APK_PERMISSIONS = "apk_permissions"
    APK_SIGNATURE = "apk_signature"
    APK_DEX = "apk_dex"
    APK_RESOURCES = "apk_resources"
    APK_NATIVE_LIBS = "apk_native_libs"
    APK_INCONSISTENCY_TESTS = "apk_inconsistency_tests"
```

---

## 6. Extractor Implementation Details

### 6.1 APKFeaturesExtractor

**File:** `redb/extractors/apk_extractors/apk_features.py`

```python
import inspect
import zipfile
from datetime import datetime, timezone
from typing import Any

from redb.extractors.enum import Tag
from redb.extractors.apk_extractor import APKExtractor
from redb.models.dataclasses import APKFeatures


class APKFeaturesExtractor(APKExtractor):

    def __init__(
        self, filepath, log, exporters=None, index_prefix=None,
        elastic_index=None, known_benign=False, known_malicious=False,
        apk=None,
    ):
        super().__init__(
            filepath, log, exporters, index_prefix,
            elastic_index, known_benign, known_malicious, apk,
        )
        self.apk_features = None
        self.elastic_index = self.index_prefix + "-apk_features"

    def tag(self):
        return Tag.APK_FEATURES.value

    def _extract_zip_timestamps(self):
        """Extract earliest and latest content modification from ZIP entries."""
        # Use zipfile directly for this — androguard doesn't expose ZIP metadata
        earliest = None
        latest = None
        try:
            zf = self._get_zip_file()
            if zf:
                with zf:
                    for info in zf.infolist():
                        try:
                            # date_time is (year, month, day, hour, minute, second)
                            dt = datetime(*info.date_time)
                            if earliest is None or dt < earliest:
                                earliest = dt
                            if latest is None or dt > latest:
                                latest = dt
                        except (ValueError, TypeError):
                            continue
        except Exception as e:
            self.log.warning(f"Error extracting ZIP timestamps: {e}")
        return (
            earliest.isoformat() if earliest else None,
            latest.isoformat() if latest else None,
        )

    def _extract_supported_abis(self):
        """Determine supported ABIs from lib/ directory."""
        abis = set()
        for f in self._list_files():
            if f.startswith("lib/") and f.endswith(".so"):
                parts = f.split("/")
                if len(parts) >= 3:
                    abis.add(parts[1])
        return sorted(abis)

    def _count_dex_files(self):
        """Count DEX files and compute total size."""
        dex_count = 0
        total_size = 0
        try:
            zf = self._get_zip_file()
            if zf:
                with zf:
                    for info in zf.infolist():
                        if info.filename.endswith(".dex"):
                            dex_count += 1
                            total_size += info.file_size
        except Exception as e:
            self.log.warning(f"Error counting DEX files: {e}")
        return dex_count, total_size

    def _check_embedded_apk(self):
        """Check if the archive contains nested APK files."""
        for f in self._list_files():
            if f.lower().endswith(".apk"):
                return True
        return False

    def extract(self):
        try:
            if not self._is_valid_apk():
                self.log.error(f"Invalid APK for {self.hash.sha256}")
                return None

            # Core metadata from androguard
            package_name = self.apk.get_package() or ""
            app_name = self.apk.get_app_name() or ""

            # version_code may be string or int
            try:
                version_code = int(self.apk.get_androidversion_code() or 0)
            except (ValueError, TypeError):
                version_code = 0

            version_name = self.apk.get_androidversion_name() or ""

            # SDK versions
            min_sdk = self.apk.get_min_sdk_version()
            target_sdk = self.apk.get_target_sdk_version()
            # NOTE: compile_sdk may not be available in all androguard versions
            compile_sdk = None
            try:
                compile_sdk = self.apk.get_effective_target_sdk_version()
            except AttributeError:
                pass

            min_sdk = int(min_sdk) if min_sdk else None
            target_sdk = int(target_sdk) if target_sdk else None
            compile_sdk = int(compile_sdk) if compile_sdk else None

            # Main activity
            main_activity = self.apk.get_main_activity()

            # Flags from manifest
            # androguard provides get_attribute_value for manifest attributes
            is_debuggable = self.apk.get_attribute_value(
                "application", "debuggable"
            ) == "true"
            allow_backup = self.apk.get_attribute_value(
                "application", "allowBackup"
            ) != "false"  # default is true
            uses_cleartext = self.apk.get_attribute_value(
                "application", "usesCleartextTraffic"
            ) == "true"

            # Archive-level analysis
            supported_abis = self._extract_supported_abis()
            dex_count, total_dex_size = self._count_dex_files()
            all_files = self._list_files()
            total_file_count = len(all_files)
            has_native_code = any(
                f.startswith("lib/") and f.endswith(".so") for f in all_files
            )
            has_assets = any(f.startswith("assets/") for f in all_files)
            contains_embedded_apk = self._check_embedded_apk()

            # Uses-library
            uses_libraries = []
            try:
                libs = self.apk.get_libraries()
                uses_libraries = list(libs) if libs else []
            except Exception:
                pass

            # ZIP timestamps
            earliest_mod, latest_mod = self._extract_zip_timestamps()

            self.apk_features = APKFeatures(
                package_name=package_name,
                app_name=app_name,
                version_code=version_code,
                version_name=version_name,
                min_sdk_version=min_sdk,
                target_sdk_version=target_sdk,
                compile_sdk_version=compile_sdk,
                main_activity=main_activity,
                is_debuggable=is_debuggable,
                allow_backup=allow_backup,
                uses_cleartext_traffic=uses_cleartext,
                supported_abis=supported_abis,
                dex_count=dex_count,
                total_dex_size=total_dex_size,
                total_file_count=total_file_count,
                has_native_code=has_native_code,
                has_assets=has_assets,
                uses_libraries=uses_libraries,
                earliest_content_modification=earliest_mod,
                latest_content_modification=latest_mod,
                contains_embedded_apk=contains_embedded_apk,
            )
            return self.apk_features

        except Exception as e:
            self.log.error(
                f"Error extracting APK features {self.hash.sha256}: {e}"
            )
            return None

    def prepare_export_data(self, exporter_type: str) -> Any:
        if exporter_type == "ElasticsearchExporter":
            return self.apk_features
        elif exporter_type == "ClickHouseExporter":
            if not self.apk_features:
                return None

            f = self.apk_features
            data = [[
                self.sha256, self.md5, self.sha1,
                f.package_name,
                f.app_name,
                f.version_code,
                f.version_name,
                f.min_sdk_version,
                f.target_sdk_version,
                f.compile_sdk_version,
                f.main_activity,
                int(f.is_debuggable),
                int(f.allow_backup),
                int(f.uses_cleartext_traffic),
                f.supported_abis,
                f.dex_count,
                f.total_dex_size,
                f.total_file_count,
                int(f.has_native_code),
                int(f.has_assets),
                f.uses_libraries,
                f.earliest_content_modification,
                f.latest_content_modification,
                int(f.contains_embedded_apk),
                datetime.now(timezone.utc),
            ]]

            column_names = [
                "sha256", "md5", "sha1",
                "package_name", "app_name", "version_code", "version_name",
                "min_sdk_version", "target_sdk_version", "compile_sdk_version",
                "main_activity",
                "is_debuggable", "allow_backup", "uses_cleartext_traffic",
                "supported_abis",
                "dex_count", "total_dex_size", "total_file_count",
                "has_native_code", "has_assets",
                "uses_libraries",
                "earliest_content_modification", "latest_content_modification",
                "contains_embedded_apk",
                "analysis_date",
            ]

            column_type_names = [
                "FixedString(64)", "FixedString(32)", "FixedString(40)",
                "String", "String", "UInt32", "String",
                "Nullable(UInt16)", "Nullable(UInt16)", "Nullable(UInt16)",
                "Nullable(String)",
                "UInt8", "UInt8", "UInt8",
                "Array(String)",
                "UInt16", "UInt64", "UInt32",
                "UInt8", "UInt8",
                "Array(String)",
                "Nullable(String)", "Nullable(String)",
                "UInt8",
                "DateTime64(3, 'UTC')",
            ]

            return (data, column_names, column_type_names)

    def get_clickhouse_table(self) -> str:
        return "redb_apk_features"
```

### 6.2 APKManifestExtractor

**File:** `redb/extractors/apk_extractors/apk_manifest.py`

**Key androguard API calls:**
```python
# Activities, services, receivers, providers
self.apk.get_activities()       # Returns list of activity class names
self.apk.get_services()         # Returns list of service class names
self.apk.get_receivers()        # Returns list of receiver class names
self.apk.get_providers()        # Returns list of provider class names

# Exported check (per component)
self.apk.get_intent_filters(component_type, component_name)
# Returns dict: {"action": [...], "category": [...]}

# Uses-features
self.apk.get_features()         # Returns list of feature strings

# Full XML
self.apk.get_android_manifest_xml().toxml()
# OR
self.apk.get_android_manifest_axml().get_xml()
```

**Implementation notes:**
- For ClickHouse export of components, use a **one-row-per-component** table (`redb_apk_components`) with columns: `sha256`, `component_type`, `class_name`, `is_exported`, `intent_actions` (Array), `intent_categories` (Array), `analysis_date`
- For the aggregated manifest data (counts, feature list, full XML), use a **one-row-per-APK** table (`redb_apk_manifest`)
- To determine if a component is exported: check `android:exported` attribute. If not explicitly set, it's implicitly exported if the component has intent filters (Android default behavior pre-API 31). From API 31+, `android:exported` must be explicit

### 6.3 APKPermissionsExtractor

**File:** `redb/extractors/apk_extractors/apk_permissions.py`

**Key androguard API calls:**
```python
# Requested permissions
self.apk.get_permissions()              # Returns list of permission strings
self.apk.get_declared_permissions()     # Custom permissions defined by the app

# For protection level, use androguard's permission mapping:
from androguard.core.api_specific_resources import load_permission_mappings
permission_map = load_permission_mappings(target_sdk)
# permission_map[perm_name] -> {"protectionLevel": "dangerous", ...}
```

**Permhash computation:**
```python
import permhash
ph = permhash.permhash_apk(self.filepath)
# Returns SHA-256 hex string
```

**Known dangerous permissions** (reference list for classification fallback):
```python
DANGEROUS_PERMISSIONS = {
    "android.permission.READ_CALENDAR",
    "android.permission.WRITE_CALENDAR",
    "android.permission.CAMERA",
    "android.permission.READ_CONTACTS",
    "android.permission.WRITE_CONTACTS",
    "android.permission.GET_ACCOUNTS",
    "android.permission.ACCESS_FINE_LOCATION",
    "android.permission.ACCESS_COARSE_LOCATION",
    "android.permission.ACCESS_BACKGROUND_LOCATION",
    "android.permission.RECORD_AUDIO",
    "android.permission.READ_PHONE_STATE",
    "android.permission.READ_PHONE_NUMBERS",
    "android.permission.CALL_PHONE",
    "android.permission.ANSWER_PHONE_CALLS",
    "android.permission.ADD_VOICEMAIL",
    "android.permission.USE_SIP",
    "android.permission.BODY_SENSORS",
    "android.permission.SEND_SMS",
    "android.permission.RECEIVE_SMS",
    "android.permission.READ_SMS",
    "android.permission.RECEIVE_WAP_PUSH",
    "android.permission.RECEIVE_MMS",
    "android.permission.READ_EXTERNAL_STORAGE",
    "android.permission.WRITE_EXTERNAL_STORAGE",
    "android.permission.READ_MEDIA_IMAGES",
    "android.permission.READ_MEDIA_VIDEO",
    "android.permission.READ_MEDIA_AUDIO",
    "android.permission.POST_NOTIFICATIONS",
    "android.permission.NEARBY_WIFI_DEVICES",
    "android.permission.ACTIVITY_RECOGNITION",
    "android.permission.BLUETOOTH_SCAN",
    "android.permission.BLUETOOTH_ADVERTISE",
    "android.permission.BLUETOOTH_CONNECT",
}
```

**ClickHouse export:** One row per permission per APK in `redb_apk_permissions`:
- `sha256`, `permission_name`, `protection_level`, `is_custom`, `analysis_date`

The `permhash` value is also stored separately — see Section 8 for integration with `HashExtractor`.

### 6.4 APKSignatureExtractor

**File:** `redb/extractors/apk_extractors/apk_signature.py`

**Key androguard API calls:**
```python
# Get certificate objects
certs = self.apk.get_certificates()     # Returns list of asn1crypto Certificate objects

# For each certificate:
cert.subject.human_friendly              # Subject DN
cert.issuer.human_friendly               # Issuer DN
cert.serial_number                       # Serial
cert.not_valid_before                    # datetime
cert.not_valid_after                     # datetime
cert.hash_algo                           # e.g., "sha256"
cert.signature_algo                      # e.g., "rsassa_pkcs1v15"

# SHA-1 thumbprint
import hashlib
hashlib.sha1(cert.dump()).hexdigest()

# SHA-256 thumbprint
hashlib.sha256(cert.dump()).hexdigest()

# Self-signed check
cert.subject == cert.issuer

# Signature scheme detection:
# v1: Check META-INF/*.SF and META-INF/*.RSA/DSA/EC files exist
# v2/v3: androguard >= 4.x has is_signed_v2() / is_signed_v3()
self.apk.is_signed_v1()
self.apk.is_signed_v2()
self.apk.is_signed_v3()
```

**Implementation notes:**
- Mirror the structure of `PECodeSigningInfo` / `MachOCodeSigningInfo`
- The `_id` field should be `self.sha256` (same as PE/Mach-O signature extractors)

### 6.5 APKDexExtractor

**File:** `redb/extractors/apk_extractors/apk_dex.py`

This is the most complex extractor. Use `androguard.core.dex` for DEX parsing.

**Key androguard API calls:**
```python
from androguard.core.dex import DEX

# Get raw DEX data from APK
dex_files = self.apk.get_all_dex()  # Returns list of raw DEX bytes

for dex_name, dex_data in zip(self.apk.get_dex_names(), dex_files):
    d = DEX(dex_data)

    # Class enumeration
    classes = d.get_classes()
    for cls in classes:
        cls.get_name()          # e.g., "Lcom/example/MainActivity;"

    # Method enumeration
    methods = d.get_methods()
    for method in methods:
        method.get_class_name()  # e.g., "Lcom/example/MainActivity;"
        method.get_name()        # e.g., "onCreate"
        method.get_descriptor()  # e.g., "(Landroid/os/Bundle;)V"

    # String constants
    strings = d.get_strings()    # Returns list of string constants
```

**Computing `top_packages`:**
```python
from collections import Counter

package_counter = Counter()
for cls in classes:
    name = cls.get_name()  # "Lcom/example/foo/Bar;"
    # Convert to package: "com.example.foo"
    parts = name[1:].replace("/", ".").rsplit(".", 1)
    if len(parts) > 1:
        package = parts[0]
    else:
        package = "(default)"
    package_counter[package] += 1

top_packages = [
    {"package": pkg, "class_count": count}
    for pkg, count in package_counter.most_common(20)
]
```

**Computing `api_usage`:**
Iterate over all method cross-references (xrefs) or method invocations. For each invoked method, check if it matches any sensitive API pattern. See Section 12 for the full categorization reference.

```python
api_usage = {cat: [] for cat in SENSITIVE_API_CATEGORIES}

for method in d.get_methods():
    # Get the bytecode and look for invoke-* instructions
    code = method.get_code()
    if not code:
        continue
    # Use androguard's analysis or instruction iteration
    # to find method references that match sensitive patterns
```

**Computing `obfuscation_indicators`:**
```python
class_names = [cls.get_name() for cls in classes]
method_names = [m.get_name() for m in methods if m.get_name() not in ("<init>", "<clinit>")]

short_class = sum(1 for n in class_names if len(n.split("/")[-1].rstrip(";")) <= 2)
short_method = sum(1 for n in method_names if len(n) <= 2)

obfuscation_indicators = {
    "short_class_names_pct": round(short_class / max(len(class_names), 1), 4),
    "short_method_names_pct": round(short_method / max(len(method_names), 1), 4),
    "non_ascii_identifiers": sum(1 for n in class_names if not n.isascii()),
    "avg_class_name_length": round(
        sum(len(n) for n in class_names) / max(len(class_names), 1), 2
    ),
}
```

**ClickHouse export:**
- `redb_apk_dex`: One row per DEX file — `sha256`, `dex_sha256`, `dex_filename`, `class_count`, `method_count`, `string_count`, `top_packages` (JSON string), `obfuscation_indicators` (JSON string), `analysis_date`
- `redb_apk_dex_api_usage`: One row per API category per DEX — `sha256`, `dex_sha256`, `category`, `api_calls` (Array(String)), `analysis_date`

### 6.6 APKResourceExtractor

**File:** `redb/extractors/apk_extractors/apk_resources.py`

**Implementation approach:**
```python
import hashlib
from magika import Magika

magika = Magika()

resources = []
suspicious_files = []

zf = self._get_zip_file()
if zf:
    with zf:
        for info in zf.infolist():
            if info.is_dir():
                continue
            # Only scan res/ and assets/ directories
            if not (info.filename.startswith("res/") or
                    info.filename.startswith("assets/")):
                continue

            data = zf.read(info.filename)
            file_sha256 = hashlib.sha256(data).hexdigest()
            filetype = magika.identify_bytes(data).output.label

            resource = APKResource(
                path=info.filename,
                size=info.file_size,
                sha256=file_sha256,
                filetype_magika=filetype,
            )

            # Flag suspicious file types
            SUSPICIOUS_TYPES = {
                "elf", "pebin", "dex", "apk", "zip", "shell",
                "javascript", "python", "powershell", "batch",
            }
            if filetype in SUSPICIOUS_TYPES:
                resource.is_suspicious = True
                suspicious_files.append(resource)

            resources.append(resource)
```

**ClickHouse export:** One row per resource in `redb_apk_resources`:
- `sha256`, `resource_path`, `resource_size`, `resource_sha256`, `filetype_magika`, `is_suspicious`, `analysis_date`

**Performance note:** For APKs with very large resource directories (some apps have 10K+ files in `res/`), consider limiting to the top N resources by size or only scanning `assets/` fully while sampling `res/`. Document any limits applied.

### 6.7 APKNativeLibExtractor

**File:** `redb/extractors/apk_extractors/apk_native_libs.py`

**Implementation approach:**
```python
native_libs = []
abis = set()

zf = self._get_zip_file()
if zf:
    with zf:
        for info in zf.infolist():
            if info.filename.startswith("lib/") and info.filename.endswith(".so"):
                parts = info.filename.split("/")
                if len(parts) >= 3:
                    abi = parts[1]
                    filename = parts[-1]
                    abis.add(abi)

                    data = zf.read(info.filename)
                    lib_sha256 = hashlib.sha256(data).hexdigest()

                    native_libs.append(APKNativeLib(
                        abi=abi,
                        filename=filename,
                        size=info.file_size,
                        sha256=lib_sha256,
                        is_known_packer=filename in KNOWN_PACKER_LIBS,
                    ))
```

See Section 11 for the `KNOWN_PACKER_LIBS` reference list.

### 6.8 APKInconsistencyTestsExtractor

**File:** `redb/extractors/apk_extractors/apk_inconsistency_tests.py`

Each test is a private method returning `bool`:

```python
def _test_zip_bomb(self):
    """Check if any ZIP entry has compression ratio > 100:1."""
    zf = self._get_zip_file()
    if not zf:
        return None
    with zf:
        for info in zf.infolist():
            if info.compress_size > 0:
                ratio = info.file_size / info.compress_size
                if ratio > 100:
                    return True
    return False

def _test_zip_duplicate_entries(self):
    """Check for duplicate filenames in ZIP directory."""
    zf = self._get_zip_file()
    if not zf:
        return None
    with zf:
        names = [info.filename for info in zf.infolist()]
        return len(names) != len(set(names))

def _test_zip_path_traversal(self):
    """Check for path traversal (../) in ZIP entry names."""
    for f in self._list_files():
        if ".." in f or f.startswith("/"):
            return True
    return False

def _test_zip_suspicious_timestamps(self):
    """Check for timestamps at epoch (1980) or in the future."""
    zf = self._get_zip_file()
    if not zf:
        return None
    now = datetime.now()
    with zf:
        for info in zf.infolist():
            try:
                dt = datetime(*info.date_time)
                if dt.year <= 1980 or dt > now:
                    return True
            except (ValueError, TypeError):
                continue
    return False

def _test_hidden_dex_files(self):
    """Check for DEX files not matching classes*.dex pattern."""
    import re
    standard_pattern = re.compile(r"^classes\d*\.dex$")
    for f in self._list_files():
        if f.endswith(".dex") and not standard_pattern.match(f.split("/")[-1]):
            return True
    return False

def _test_emulator_detection_strings(self):
    """Check for emulator detection patterns in DEX strings."""
    # Inspect DEX strings for known emulator detection indicators
    EMULATOR_INDICATORS = {
        "generic", "sdk", "google_sdk", "Emulator",
        "goldfish", "ranchu", "Andy", "Genymotion",
        "BlueStacks", "nox", "ttVM_Hdragon",
    }
    # Use androguard to extract strings from DEX, then check
    ...

def _test_debugger_detection(self):
    """Check for debugger detection API calls in DEX."""
    # Look for: Debug.isDebuggerConnected, Debug.waitingForDebugger
    ...

def _test_root_detection(self):
    """Check for root detection patterns in DEX."""
    ROOT_INDICATORS = {
        "/system/app/Superuser.apk",
        "/system/xbin/su",
        "/system/bin/su",
        "com.noshufou.android.su",
        "com.thirdparty.superuser",
        "eu.chainfire.supersu",
        "com.koushikdutta.superuser",
        "com.topjohnwu.magisk",
    }
    ...
```

---

## 7. Ingestor Integration

**File:** `redb/ingestor.py`

Replace the stub at line 1668:

```python
elif filetype == "apk":
    logger.debug("APK file detected")
```

With the full dispatch block:

```python
elif filetype == "apk":
    logger.debug("APK file detected")

    # Import APK extractors
    from redb.extractors.apk_extractor import APKExtractor
    from redb.extractors.apk_extractors.apk_features import APKFeaturesExtractor
    from redb.extractors.apk_extractors.apk_manifest import APKManifestExtractor
    from redb.extractors.apk_extractors.apk_permissions import APKPermissionsExtractor
    from redb.extractors.apk_extractors.apk_signature import APKSignatureExtractor
    from redb.extractors.apk_extractors.apk_dex import APKDexExtractor
    from redb.extractors.apk_extractors.apk_resources import APKResourceExtractor
    from redb.extractors.apk_extractors.apk_native_libs import APKNativeLibExtractor
    from redb.extractors.apk_extractors.apk_inconsistency_tests import APKInconsistencyTestsExtractor

    apk_modules = [
        APKFeaturesExtractor,
        APKManifestExtractor,
        APKPermissionsExtractor,
        APKSignatureExtractor,
        APKDexExtractor,
        APKResourceExtractor,
        APKNativeLibExtractor,
        APKInconsistencyTestsExtractor,
    ]

    # Parse APK once with androguard, share across all extractors
    try:
        from androguard.core.apk import APK
        apk = APK(filepath)
    except Exception as e:
        logger.error(f"Failed to parse APK {filepath}: {e}")
        apk = None

    # Run format-agnostic extractors first (same as PE/ELF/MachO)
    # DIE
    if 'all' in selected_modules or 'DIEExtractor' in selected_modules:
        try:
            extractor = DIEExtractor(
                filepath, logger, exporters=exporters,
                index_prefix=index_prefix
            )
            result = extractor.export_data()
            is_packed = bool(result) if result is not None else None
            if result:
                results[ImportResult.CORRECTLY] += 1
            elif result is False:
                results[ImportResult.FAILED] += 1
        except Exception as e:
            logger.error(f"Error in DIEExtractor: {str(e)}")
            results[ImportResult.FAILED] += 1

    # BasicProperties
    if 'all' in selected_modules or 'BasicPropertiesExtractor' in selected_modules:
        try:
            extractor = BasicPropertiesExtractor(
                filepath, logger, exporters=exporters,
                index_prefix=index_prefix
            )
            if is_packed is not None:
                extractor.is_packed = is_packed
            result = extractor.export_data()
            if result:
                results[ImportResult.CORRECTLY] += 1
            elif result is False:
                results[ImportResult.FAILED] += 1
        except Exception as e:
            logger.error(f"Error in BasicPropertiesExtractor: {str(e)}")
            results[ImportResult.FAILED] += 1

    # Hashes
    if 'all' in selected_modules or 'HashExtractor' in selected_modules:
        try:
            extractor = HashExtractor(
                filepath, logger, exporters=exporters,
                index_prefix=index_prefix
            )
            result = extractor.export_data()
            if result:
                results[ImportResult.CORRECTLY] += 1
            elif result is False:
                results[ImportResult.FAILED] += 1
        except Exception as e:
            logger.error(f"Error in HashExtractor: {str(e)}")
            results[ImportResult.FAILED] += 1

    # YARA
    if 'all' in selected_modules or 'YaraExtractor' in selected_modules:
        try:
            extractor = YaraExtractor(
                filepath, logger, exporters=exporters,
                index_prefix=index_prefix
            )
            result = extractor.export_data()
            if result:
                results[ImportResult.CORRECTLY] += 1
            elif result is False:
                results[ImportResult.FAILED] += 1
        except Exception as e:
            logger.error(f"Error in YaraExtractor: {str(e)}")
            results[ImportResult.FAILED] += 1

    # APK-specific extractors
    for module in apk_modules:
        if 'all' in selected_modules or module.__name__ in selected_modules:
            try:
                logger.debug(f"Running {module.__name__}")
                extractor = module(
                    filepath,
                    logger,
                    exporters=exporters,
                    index_prefix=index_prefix,
                    apk=apk
                )
                result = extractor.export_data()
                if result:
                    results[ImportResult.CORRECTLY] += 1
                elif result is False:
                    results[ImportResult.FAILED] += 1
            except Exception as e:
                logger.error(f"Error in {module.__name__}: {str(e)}")
                results[ImportResult.FAILED] += 1
```

**Note:** Follow the exact error-handling and `selected_modules` check pattern used by the existing PE/ELF/Mach-O dispatch blocks. The code above mirrors the pattern from `ingestor.py:1540-1666`.

---

## 8. Hashes Dataclass Update

**File:** `redb/models/dataclasses.py`

Add to the `Hashes` dataclass after the existing Mach-O hashes:

```python
@dataclass
class Hashes:
    md5: str
    sha1: str
    sha256: str
    ssdeep_hash: str
    tlsh_hash: str
    authentihash: Optional[str] = None  # PE
    imphash: Optional[str] = None  # PE
    impfuzzy: Optional[str] = None  # PE TODO: implement
    typerefhash: Optional[str] = None  # DotNet
    richhash: Optional[str] = None  # PE
    richpe_hash: Optional[str] = None  # PE
    richpv_hash: Optional[str] = None  # PE
    richpv_hash_sorted: Optional[str] = None  # PE
    import_hash: Optional[str] = None  # ELF
    export_hash: Optional[str] = None  # ELF
    section_hash: Optional[str] = None  # ELF
    symbol_hash: Optional[str] = None  # ELF (symhash)
    dynamic_hash: Optional[str] = None  # ELF
    # Mach-O similarity hashes
    macho_dylib_hash: Optional[str] = None
    macho_import_hash: Optional[str] = None
    macho_export_hash: Optional[str] = None
    macho_entitlement_hash: Optional[str] = None
    macho_symhash: Optional[str] = None
    # APK similarity hashes
    permhash: Optional[str] = None  # APK: SHA-256 of sorted permissions (Mandiant/Google)
```

The `permhash` should be computed in `HashExtractor` when the filetype is APK, using the `permhash` library:

```python
# In HashExtractor, when filetype == "apk":
import permhash as permhash_lib
self.hashes.permhash = permhash_lib.permhash_apk(self.filepath)
```

The `permhash` column must be added to the existing `redb_hashes` table:

```sql
ALTER TABLE redb_hashes
    ADD COLUMN IF NOT EXISTS permhash Nullable(FixedString(64));
```

---

## 9. ClickHouse Schema

Create the following tables. Naming convention follows existing tables (`redb_pe_features`, `redb_elf_features`, etc.):

```sql
-- Core metadata (one row per APK)
CREATE TABLE redb_apk_features (
    sha256              FixedString(64),
    md5                 FixedString(32),
    sha1                FixedString(40),
    package_name        String,
    app_name            String,
    version_code        UInt32,
    version_name        String,
    min_sdk_version     Nullable(UInt16),
    target_sdk_version  Nullable(UInt16),
    compile_sdk_version Nullable(UInt16),
    main_activity       Nullable(String),
    is_debuggable       UInt8,
    allow_backup        UInt8,
    uses_cleartext_traffic UInt8,
    supported_abis      Array(String),
    dex_count           UInt16,
    total_dex_size      UInt64,
    total_file_count    UInt32,
    has_native_code     UInt8,
    has_assets          UInt8,
    uses_libraries      Array(String),
    earliest_content_modification Nullable(String),
    latest_content_modification   Nullable(String),
    contains_embedded_apk UInt8,
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY sha256;

-- Manifest components (one row per component)
CREATE TABLE redb_apk_components (
    sha256              FixedString(64),
    component_type      LowCardinality(String),  -- "activity", "service", "receiver", "provider"
    class_name          String,
    is_exported         UInt8,
    intent_actions      Array(String),
    intent_categories   Array(String),
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY (sha256, component_type, class_name);

-- Manifest summary (one row per APK)
CREATE TABLE redb_apk_manifest (
    sha256              FixedString(64),
    md5                 FixedString(32),
    sha1                FixedString(40),
    activity_count      UInt16,
    service_count       UInt16,
    receiver_count      UInt16,
    provider_count      UInt16,
    intent_filters_by_action   Array(String),
    intent_filters_by_category Array(String),
    uses_features       Array(String),
    manifest_xml        Nullable(String),
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY sha256;

-- Permissions (one row per permission per APK)
CREATE TABLE redb_apk_permissions (
    sha256              FixedString(64),
    permission_name     String,
    protection_level    LowCardinality(String),
    is_custom           UInt8,
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY (sha256, permission_name);

-- Signing info (one row per APK)
CREATE TABLE redb_apk_signature (
    sha256              FixedString(64),
    md5                 FixedString(32),
    sha1                FixedString(40),
    is_signed           UInt8,
    signature_scheme_versions Array(UInt8),
    number_of_certificates UInt8,
    signer_subject      Nullable(String),
    signer_issuer       Nullable(String),
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY sha256;

-- Certificate details (one row per certificate per APK)
CREATE TABLE redb_apk_certificates (
    sha256              FixedString(64),
    subject             String,
    issuer              String,
    serial_number       String,
    valid_from          String,
    valid_to            String,
    thumbprint_sha1     FixedString(40),
    thumbprint_sha256   FixedString(64),
    algorithm           Nullable(String),
    key_size            Nullable(UInt16),
    is_self_signed      UInt8,
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY (sha256, thumbprint_sha256);

-- DEX file analysis (one row per DEX file)
CREATE TABLE redb_apk_dex (
    sha256              FixedString(64),
    dex_sha256          FixedString(64),
    dex_filename        String,
    dex_tlsh            Nullable(String),
    class_count         UInt32,
    method_count        UInt32,
    string_count        UInt32,
    top_packages        String,       -- JSON
    obfuscation_indicators String,    -- JSON
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY (sha256, dex_sha256);

-- DEX API usage (one row per category per DEX)
CREATE TABLE redb_apk_dex_api_usage (
    sha256              FixedString(64),
    dex_sha256          FixedString(64),
    category            LowCardinality(String),
    api_calls           Array(String),
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY (sha256, dex_sha256, category);

-- Resource inventory (one row per resource)
CREATE TABLE redb_apk_resources (
    sha256              FixedString(64),
    resource_path       String,
    resource_size       UInt64,
    resource_sha256     FixedString(64),
    filetype_magika     LowCardinality(String),
    is_suspicious       UInt8,
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY (sha256, resource_path);

-- Native libraries (one row per .so file)
CREATE TABLE redb_apk_native_libs (
    sha256              FixedString(64),
    abi                 LowCardinality(String),
    filename            String,
    size                UInt64,
    lib_sha256          FixedString(64),
    is_known_packer     UInt8,
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY (sha256, abi, filename);

-- Inconsistency tests (one row per APK)
CREATE TABLE redb_apk_inconsistency_tests (
    sha256              FixedString(64),
    md5                 FixedString(32),
    sha1                FixedString(40),
    test_zip_bomb                    Nullable(UInt8),
    test_zip_duplicate_entries       Nullable(UInt8),
    test_zip_path_traversal          Nullable(UInt8),
    test_zip_suspicious_timestamps   Nullable(UInt8),
    test_hidden_dex_files            Nullable(UInt8),
    test_manifest_component_mismatch Nullable(UInt8),
    test_debuggable_release          Nullable(UInt8),
    test_emulator_detection_strings  Nullable(UInt8),
    test_debugger_detection          Nullable(UInt8),
    test_root_detection              Nullable(UInt8),
    analysis_date       DateTime64(3, 'UTC')
) ENGINE = ReplacingMergeTree()
ORDER BY sha256;
```

---

## 10. Androguard API Reference

Quick reference for the most commonly used androguard APIs. Full docs: https://androguard.readthedocs.io/

### APK class (`androguard.core.apk.APK`)

| Method | Returns | Description |
|--------|---------|-------------|
| `APK(filepath)` | `APK` | Parse an APK file |
| `.is_valid_APK()` | `bool` | Check if parsing succeeded |
| `.get_package()` | `str` | Package name |
| `.get_app_name()` | `str` | Application display name |
| `.get_androidversion_code()` | `str` | Version code |
| `.get_androidversion_name()` | `str` | Version name |
| `.get_min_sdk_version()` | `str` | Minimum SDK |
| `.get_target_sdk_version()` | `str` | Target SDK |
| `.get_main_activity()` | `str` | Main/launcher activity |
| `.get_activities()` | `list[str]` | All activity class names |
| `.get_services()` | `list[str]` | All service class names |
| `.get_receivers()` | `list[str]` | All receiver class names |
| `.get_providers()` | `list[str]` | All provider class names |
| `.get_permissions()` | `list[str]` | Requested permissions |
| `.get_declared_permissions()` | `list[str]` | Custom-defined permissions |
| `.get_features()` | `list[str]` | Uses-feature declarations |
| `.get_libraries()` | `list[str]` | Uses-library declarations |
| `.get_files()` | `list[str]` | All files in the APK |
| `.get_certificates()` | `list[Certificate]` | X.509 certificate objects |
| `.get_all_dex()` | `list[bytes]` | Raw DEX file contents |
| `.get_dex_names()` | `list[str]` | DEX filenames |
| `.is_signed_v1()` | `bool` | JAR signature present |
| `.is_signed_v2()` | `bool` | v2 signature present |
| `.is_signed_v3()` | `bool` | v3 signature present |
| `.get_attribute_value(tag, attr)` | `str` | Get manifest attribute |
| `.get_android_manifest_xml()` | `minidom` | Parsed manifest XML |
| `.get_intent_filters(type, name)` | `dict` | Intent filters for component |

### DEX class (`androguard.core.dex.DEX`)

| Method | Returns | Description |
|--------|---------|-------------|
| `DEX(raw_bytes)` | `DEX` | Parse raw DEX data |
| `.get_classes()` | `list[ClassDefItem]` | All class definitions |
| `.get_methods()` | `list[EncodedMethod]` | All methods |
| `.get_strings()` | `list[str]` | All string constants |

---

## 11. Known Packer/Protector Signatures

Native library names commonly associated with packers and protectors. Used by `APKNativeLibExtractor` to flag `is_known_packer`:

```python
KNOWN_PACKER_LIBS = {
    # Jiagu (360/Qihoo)
    "libjiagu.so", "libjiagu_a64.so", "libjiagu_x86.so", "libjiagu_x64.so",
    # Bangcle/SecNeo
    "libsecexe.so", "libsecmain.so", "libSecShell.so",
    # Baidu
    "libbaiduprotect.so",
    # Tencent (Legu)
    "libshell-super.2019.so", "libshella-*.so", "libtxAppProtect.so",
    "libBugly.so", "mix.dex",
    # iJiami
    "libexec.so", "libexecmain.so",
    # Alibaba
    "libmobisec.so", "libaliprotect.so",
    # APKProtect
    "libAPKProtect.so",
    # Pangxie (Pangolin)
    "libdexjni.so",
    # DexProtector
    "libdexprotector.so",
    # AppSolid
    "libAppSolid.so",
    # Kiwisec
    "libkwscmm.so",
    # DingXiang
    "libx3g.so",
    # NQ Shield
    "libnqshield.so",
    # Generic / other
    "libprotectClass.so",
    "libDexHelper.so",
    "libdexloader.so",
    "libfdog.so",
}
```

**Note:** This list should be maintained and expanded over time. For glob-style matching (e.g., `libshella-*.so`), use `fnmatch` instead of exact set lookup.

---

## 12. Sensitive API Categorization Reference

Used by `APKDexExtractor` to categorize API calls. Each category maps to a list of class/method patterns:

```python
SENSITIVE_API_CATEGORIES = {
    "reflection": [
        "Ljava/lang/reflect/Method;->invoke",
        "Ljava/lang/reflect/Field;->get",
        "Ljava/lang/reflect/Field;->set",
        "Ljava/lang/reflect/Constructor;->newInstance",
        "Ljava/lang/Class;->forName",
        "Ljava/lang/Class;->getMethod",
        "Ljava/lang/Class;->getDeclaredMethod",
        "Ljava/lang/Class;->getDeclaredField",
        "Ljava/lang/ClassLoader;->loadClass",
    ],
    "crypto": [
        "Ljavax/crypto/Cipher;->getInstance",
        "Ljavax/crypto/Cipher;->init",
        "Ljavax/crypto/spec/SecretKeySpec;-><init>",
        "Ljavax/crypto/spec/IvParameterSpec;-><init>",
        "Ljava/security/MessageDigest;->getInstance",
        "Ljava/security/KeyStore;->getInstance",
        "Ljavax/crypto/Mac;->getInstance",
    ],
    "dynamic_loading": [
        "Ldalvik/system/DexClassLoader;-><init>",
        "Ldalvik/system/PathClassLoader;-><init>",
        "Ldalvik/system/InMemoryDexClassLoader;-><init>",
        "Ldalvik/system/BaseDexClassLoader;-><init>",
        "Ljava/lang/Runtime;->exec",
        "Ljava/lang/ProcessBuilder;->start",
    ],
    "telephony": [
        "Landroid/telephony/TelephonyManager;->getDeviceId",
        "Landroid/telephony/TelephonyManager;->getSubscriberId",
        "Landroid/telephony/TelephonyManager;->getLine1Number",
        "Landroid/telephony/TelephonyManager;->getSimSerialNumber",
        "Landroid/telephony/TelephonyManager;->getNetworkOperator",
        "Landroid/telephony/TelephonyManager;->getSimOperator",
    ],
    "sms": [
        "Landroid/telephony/SmsManager;->sendTextMessage",
        "Landroid/telephony/SmsManager;->sendMultipartTextMessage",
        "Landroid/telephony/SmsManager;->sendDataMessage",
    ],
    "network": [
        "Ljava/net/HttpURLConnection;->connect",
        "Ljava/net/URL;->openConnection",
        "Lokhttp3/OkHttpClient;-><init>",
        "Lokhttp3/Request$Builder;->build",
        "Lorg/apache/http/client/HttpClient;->execute",
        "Landroid/webkit/WebView;->loadUrl",
        "Landroid/webkit/WebView;->setWebViewClient",
    ],
    "native": [
        "Ljava/lang/System;->loadLibrary",
        "Ljava/lang/System;->load",
        "Ljava/lang/Runtime;->loadLibrary",
    ],
    "device_info": [
        "Landroid/os/Build;->FINGERPRINT",
        "Landroid/os/Build;->MODEL",
        "Landroid/os/Build;->MANUFACTURER",
        "Landroid/os/Build;->PRODUCT",
        "Landroid/os/Build;->BRAND",
        "Landroid/os/Build;->DEVICE",
        "Landroid/os/Build;->HARDWARE",
        "Landroid/os/Build;->SERIAL",
        "Landroid/os/Build$VERSION;->SDK_INT",
        "Landroid/provider/Settings$Secure;->getString",  # ANDROID_ID
    ],
    "file_io": [
        "Ljava/io/FileOutputStream;-><init>",
        "Ljava/io/FileInputStream;-><init>",
        "Landroid/content/SharedPreferences;->edit",
        "Landroid/database/sqlite/SQLiteDatabase;->execSQL",
        "Landroid/database/sqlite/SQLiteDatabase;->rawQuery",
    ],
    "ipc": [
        "Landroid/content/ContentResolver;->query",
        "Landroid/content/ContentResolver;->insert",
        "Landroid/content/ContentResolver;->delete",
        "Landroid/content/Intent;-><init>",
        "Landroid/content/Context;->sendBroadcast",
        "Landroid/content/Context;->startService",
        "Landroid/content/Context;->bindService",
    ],
}
```

**Matching approach:** For each method invocation found in DEX bytecode, check if its full signature (class + method) starts with any of the patterns above. Store only the first match per unique API call (avoid duplicates from multiple call sites).

---

## Implementation Checklist

- [ ] Add `androguard` and `permhash` to `requirements.txt`
- [ ] Create `redb/extractors/apk_extractor.py` (base class)
- [ ] Create `redb/extractors/apk_extractors/__init__.py`
- [ ] Implement `APKFeaturesExtractor`
- [ ] Implement `APKManifestExtractor`
- [ ] Implement `APKPermissionsExtractor`
- [ ] Implement `APKSignatureExtractor`
- [ ] Implement `APKDexExtractor`
- [ ] Implement `APKResourceExtractor`
- [ ] Implement `APKNativeLibExtractor`
- [ ] Implement `APKInconsistencyTestsExtractor`
- [ ] Add APK tags to `redb/extractors/enum.py`
- [ ] Add APK dataclasses to `redb/models/dataclasses.py`
- [ ] Add `permhash` field to `Hashes` dataclass
- [ ] Wire up APK dispatch in `redb/ingestor.py`
- [ ] Integrate `permhash` computation in `HashExtractor`
- [ ] Create ClickHouse migration scripts for new tables
- [ ] Unit tests per extractor
- [ ] Integration test: full APK ingestion pipeline
- [ ] Cross-validate output against VirusTotal for reference samples