The Agent Skills Directory

[SAFE]: The skill instructions and associated Python scripts do not contain malicious instructions, obfuscation, or unauthorized access patterns.
[COMMAND_EXECUTION]: The skill documentation provides examples for local execution of the triage tool. The script scripts/incident_triage.py is limited to data analysis and reporting, with no capabilities for arbitrary command execution or system modification.
[INDIRECT_PROMPT_INJECTION]: The triage script accepts external JSON event data, which presents a surface for indirect prompt injection.
Ingestion points: Event data is ingested in scripts/incident_triage.py using json.load() from files or standard input.
Boundary markers: None are present; the script parses the raw JSON structure.
Capability inventory: The script lacks network, file system write, and subprocess execution capabilities, preventing potential exploitation from leading to exfiltration or remote code execution.
Sanitization: Standard JSON parsing is used to ensure the input is handled as structured data rather than executable instructions.

incident-response