reorder
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [SAFE]: No malicious patterns, unauthorized data access, or suspicious external dependencies were found in the skill. The logic is restricted to non-destructive text manipulation.
- [NO_CODE]: The skill consists entirely of instructional markdown and does not ship with any executable scripts or binaries.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input in the form of listings, tables, or catalogs, creating a surface for indirect prompt injection. * Ingestion points: User-provided content targeted for reordering as described in SKILL.md. * Boundary markers: The prompt includes strict negative constraints ("Reorder only", "Reword nothing", "Add and remove nothing") which serve as effective boundaries against the execution of instructions embedded within the processed data. * Capability inventory: Text manipulation and the potential use of scripts for structural moves. * Sanitization: The workflow mandates item existence verification ("report a MISS, never a silent drop") and unicode-safe mutation.
Audit Metadata