The Agent Skills Directory

[COMMAND_EXECUTION]: The script scripts/agent.py uses subprocess.run to invoke external system utilities like photorec and file. These calls use list-based arguments which mitigate shell injection risks. The script also includes a hashing function that reads files into memory, which could lead to resource exhaustion if used on extremely large recovered files.
[PRIVILEGE_ESCALATION]: The SKILL.md file documents the use of sudo for tool installation and raw disk access. These elevated permissions are standard requirements for forensic data recovery operations.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection via the processing of untrusted forensic images.
Ingestion points: Filenames and file signatures are read from recovered files in scripts/agent.py and displayed in the report.
Boundary markers: None are present to delimit untrusted metadata from instructions.
Capability inventory: The script can execute subprocesses and modify the file system.
Sanitization: There is no sanitization of metadata recovered from the disk images.

recovering-deleted-files-with-photorec