The Agent Skills Directory

[SAFE]: The skill is entirely instructional, providing character limits and best practices for SEO meta descriptions. No malicious patterns, command executions, or remote code downloads were detected.
[INDIRECT_PROMPT_INJECTION]: The skill encourages the agent to read project context files (.claude/project-context.md, .cursor/project-context.md) to inform its suggestions.
Ingestion points: Initial Assessment section in SKILL.md directs the agent to read local project context files.
Boundary markers: Absent.
Capability inventory: No subprocess calls, network operations, or file-write capabilities were identified within this skill.
Sanitization: Absent.
Assessment: Because the skill lacks dangerous capabilities to be exploited, the use of external context files is considered a standard feature rather than a security risk.

meta-description