audit-report

Pass

Audited by Gen Agent Trust Hub on Sep 18, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill uses the harness-mcp-v2 server and authorized tools (harness_list, harness_describe) to perform audit log retrieval, which is the primary and legitimate function of the skill for the Harness platform.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted audit event data, creating a potential surface for indirect prompt injection. Ingestion points: Audit event records retrieved via the harness_list tool as described in SKILL.md. Boundary markers: No specific delimiters or instructions to ignore embedded content are provided for the log data. Capability inventory: The skill is limited to listing and describing audit events; no dangerous capabilities like subprocess execution, file writes, or unauthorized network operations were detected. Sanitization: No explicit sanitization or filtering of audit event content is performed before interpolation into the report templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 18, 2026, 11:25 PM
Security Audit — agent-trust-hub — audit-report