The Agent Skills Directory

[SAFE]: The skill defines an analytical framework (ABLE) and scoring system for threat hunting. It contains no executable code, network operations, or credential access. All logic is focused on documentation and methodology.
[INDIRECT_PROMPT_INJECTION]: The skill identifies an attack surface for indirect prompt injection as it ingests external "source intelligence" (Phase 1, Step 1). 1. Ingestion points: "source intelligence" (SKILL.md). 2. Boundary markers: Absent in the rationale template. 3. Capability inventory: The skill itself has no executable capabilities, but it guides query creation for external SIEM/EDR platforms. 4. Sanitization: No explicit sanitization or filtering of intelligence content is mandated beyond verbatim quoting. As this ingestion is essential to the threat hunting process, the severity is assessed as safe.

threat-hunting