The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts scripts/agent.py and scripts/process.py use subprocess.run to execute external security and system tools like ldapsearch, wevtutil, and the impacket library. This functionality is core to the skill's purpose but involves running commands based on user-provided parameters.
[CREDENTIALS_UNSAFE]: The scripts/process.py script accepts domain credentials, including plaintext passwords and NTLM hashes, via command-line arguments. This can lead to sensitive information being exposed in process listings or shell history files.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through the ingestion and analysis of external data.
Ingestion points: The functions enumerate_spn_accounts and detect_kerberoasting in scripts/agent.py and scripts/process.py read data from Active Directory (LDAP) and Windows Security Event logs (.evtx files).
Boundary markers: There are no boundary markers or instructions to ignore instructions embedded within the ingested data.
Capability inventory: The skill has the capability to execute shell commands, perform network operations (LDAP), and write files (reports), which could be leveraged if malicious instructions were successfully injected into the processed logs or directory attributes.
Sanitization: The skill lacks sanitization for the data retrieved from external sources before it is processed or incorporated into generated reports.

performing-kerberoasting-attack