write-aida-article-review
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches article review prompts and style guides from the author's domain
jeffbailey.us. These resources are used to define the evaluation logic for the agent. - [COMMAND_EXECUTION]: Uses
curlandpandocshell commands to download, convert, and cache remote content into the~/.claude/cache/writing-prompts/directory. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it processes untrusted user-provided articles and external prompt templates.
- Ingestion points: Remote rubric files (e.g.,
aida-article-review.md) and user-provided article text or files. - Boundary markers: No specific delimiters or safety instructions are defined to separate untrusted content from the system instructions.
- Capability inventory: The skill can perform network requests via
curland write/read files within its local cache directory. - Sanitization: No explicit sanitization or validation of template variables or article content is described.
Audit Metadata