review-attack-path
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill dynamically constructs Cypher queries for use with the subimageRunCypher tool. It includes explicit security instructions to escape single quotes and backslashes in user-provided identifiers (such as ARNs and asset IDs) to mitigate query injection risks.
- [DYNAMIC_EXECUTION]: The skill generates and executes database scripts (Cypher) at runtime based on user input and pre-defined templates to explore graph schemas and retrieve path details.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from an external Neo4j graph database to summarize attack paths and influence downstream actions like ticket creation or notifications. Ingestion points: Data returned from subimageRunCypher. Boundary markers: None explicitly used for output processing, though confidence tiers (Confirmed, Inferred, Hypothetical) are defined to manage data trust. Capability inventory: subimageRunCypher, subimageCreateTicket, subimageSendNotification, and subimageCreateAttackPathScenario. Sanitization: Input escaping for Cypher queries is instructed for the agent.
Audit Metadata