backlink-analyzer

Pass

Audited by Gen Agent Trust Hub on May 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious behavior or suspicious code patterns were detected. The skill instructions and referenced materials are consistent with the stated purpose of SEO auditing and backlink analysis.
  • [DATA_EXFILTRATION]: The skill records its findings in local project directories such as memory/monitoring/, which is a standard practice for maintaining agent state and does not constitute unauthorized data exfiltration.
  • [PROMPT_INJECTION]: The skill processes external data including backlink CSVs and domain URLs, which creates a theoretical surface for indirect prompt injection. 1. Ingestion points: External backlink CSV files and domain URLs (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: File-write operations to project memory directories (SKILL.md). 4. Sanitization: Absent. This represents a functional surface common to analytical tools and is considered low-risk given the lack of dangerous capabilities like shell execution.
Audit Metadata
Risk Level
SAFE
Analyzed
May 14, 2026, 11:37 AM