The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The script scripts/agent.py is designed to accept sensitive credentials, including a password, through command-line arguments via the --password flag. This practice is inherently insecure as command-line arguments are often visible to other users on the system through process monitoring utilities and may be persisted in shell history files or system logs.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes data from Active Directory, which is an external and potentially untrusted source. An attacker with control over directory objects could inject malicious payloads into attributes.\n
Ingestion points: Data is ingested from LDAP search results (e.g., sAMAccountName, cn, and member fields) in scripts/agent.py.\n
Capability inventory: The script can write files to the local disk and output results to the console.\n
Sanitization: The data retrieved from LDAP is not sanitized or validated before being included in the final JSON report.\n
Boundary markers: The output does not include delimiters or instructions to downstream agents to ignore embedded instructions within the account data.\n- [EXTERNAL_DOWNLOADS]: The skill requires the ldap3 library to be installed from the Python Package Index (PyPI). While ldap3 is a standard and well-known library for LDAP operations, the dependency is not pinned to a specific version in the documentation, which is a common security best practice.

performing-privileged-account-discovery