research-add-items
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection vulnerability surface because it processes untrusted content from web searches and an existing outline file, and has the capability to write to the local file system.
- Ingestion points: The skill reads
*/outline.yamlin Step 1 and ingests data from external web searches in Step 2. - Boundary markers: There are no boundary markers or instructions to isolate or ignore embedded prompts within the retrieved data.
- Capability inventory: The skill modifies and saves the updated content back to the local file system (
outline.yaml). - Sanitization: No data validation, escaping, or strict schema compliance checks are applied to the ingested external content prior to integration and file writing.
Audit Metadata