The Agent Skills Directory

[SAFE]: The skill is comprised solely of markdown files and evaluation data. There are no executable scripts, binaries, or automated shell commands included.
[SAFE]: No sensitive data access or exfiltration vectors were identified. The skill does not reference private directories, environment variables, or hardcoded secrets.
[SAFE]: External links and references are limited to reputable industry resources and official documentation for search engines and AI bots.
[SAFE]: No obfuscation, prompt injection, or persistence techniques were detected within the instructions or metadata.
[INDIRECT_PROMPT_INJECTION]: The skill is configured to read localized context files.
Ingestion points: .agents/product-marketing-context.md (specified in SKILL.md)
Boundary markers: Absent; the agent is simply instructed to read the file.
Capability inventory: The skill leverages external SEO tools (semrush, ahrefs, gsc, ga4) for data analysis.
Sanitization: None; the agent processes the file content as instructional context.
Assessment: This is a standard architectural pattern for skill personalization and does not constitute a malicious finding.

ai-seo