offensive-crypto-attacks
Pass
Audited by Gen Agent Trust Hub on Aug 27, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions to execute third-party CLI tools such as
padbuster,RsaCtfTool, andhashpumpto automate the decryption and forgery of cryptographic tokens. - [EXTERNAL_DOWNLOADS]: The skill recommends the installation of external Python packages like
hashpumpyandpycryptodometo perform specialized attacks like hash length extensions and custom crypto-primitive manipulation. - [INDIRECT_PROMPT_INJECTION]: The skill implements attack logic that ingests data from external HTTP endpoints (oracles) which represents a surface for untrusted data ingestion.
- Ingestion points: Network responses from target URLs and binary data from ciphertext parameters in
SKILL.md. - Boundary markers: Absent; the skill does not use delimiters to isolate external data from the execution logic.
- Capability inventory: Network requests via
requests.get, subprocess execution of offensive tools, and file system manipulation (referenced in tool usage). - Sanitization: Absent; responses are processed directly to determine oracle states.
Audit Metadata