The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an inherent attack surface because it is designed to ingest and analyze external log data (DNS queries, process event logs in JSON/CSV formats).
Ingestion points: The analyze_dns_logs function in scripts/agent.py and the parse_logs function in scripts/process.py read data from user-provided file paths.
Boundary markers: The implementation does not currently use specific delimiters or instructions to the agent to ignore potentially malicious embedded content within the logs.
Capability inventory: The Python scripts are limited to pattern matching and report generation; they do not possess dangerous capabilities like dynamic code execution or arbitrary network requests based on the ingested content.
Sanitization: Data is processed via standard JSON/CSV parsers and filtered using hardcoded regular expressions for specific cloud domains and threat indicators.
[EXTERNAL_DOWNLOADS]: The skill documentation mentions installing the elasticsearch Python library. This is a well-known, legitimate package required for interacting with official Elastic SIEM instances.
[SAFE]: Automated scan findings regarding attack.mitre.org and references/standards.md were evaluated. The URL points to the authoritative MITRE ATT&CK knowledge base, and the file contains a standard reference list of Sysmon event IDs and MITRE techniques. These are essential components for threat hunting and do not represent malicious activity.

hunting-for-living-off-the-cloud-techniques