The Agent Skills Directory

[Indirect Prompt Injection] (LOW): The skill ingests untrusted data from external files. Ingestion points: Reads essay-draft.md, essay-brief.md, and essay-outline.md. Boundary markers: None present. Capability inventory: Limited to text-based output only; no exec, eval, subprocess calls, network operations, or file-write capabilities were found. Sanitization: None. As a 'display only' skill with no side effects, the risk is minimal.
[Remote Code Execution] (SAFE): No remote code patterns or script execution commands were detected.
[Data Exfiltration] (SAFE): No network requests or sensitive file access patterns are present.

essay-review