The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses and copies sensitive Windows forensic artifacts including the SYSTEM, SOFTWARE, and NTUSER.DAT registry hives. These files contain system-wide configuration, boot keys, and user-specific activity history. The operations are performed locally within the analysis directory, but they involve the exposure of protected system data to the agent context.
[COMMAND_EXECUTION]: The skill's workflow involves executing shell commands to mount forensic images and run Python scripts for artifact parsing. This includes the 'mount' command, which typically requires root or administrative privileges to interact with loop devices and filesystem offsets.
[EXTERNAL_DOWNLOADS]: The skill depends on several third-party forensic libraries to function, specifically 'regipy', 'python-registry', and 'evtx'. These are established tools in the security community used for parsing complex Windows data structures.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it ingests and processes untrusted external data from forensic images. \n- Ingestion points: Windows registry hives (SYSTEM, SOFTWARE, NTUSER.DAT) and system event logs are read and parsed by the agent scripts. \n- Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore or isolate potentially malicious instructions embedded within the registry keys or log entries being analyzed. \n- Capability inventory: The skill is capable of local file system writes to generate JSON and CSV reports, but it does not have network access or the ability to execute arbitrary code on the host system. \n- Sanitization: The skill relies on the internal logic of third-party parsing libraries without implementing its own validation or sanitization of the content extracted from the forensic artifacts.

analyzing-usb-device-connection-history