create-prompt
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill documents a template system in
SKILL.mdthat interpolates untrusted task data into agent prompts.\n - Ingestion points: External task identifiers are ingested through
{{ticket}}variables defined inSKILL.md.\n - Boundary markers:
SKILL.mddoes not mention the use of delimiters or instructions to ignore embedded commands within the interpolated data.\n - Capability inventory: The examples in
SKILL.mddescribe an agent capable of modifying tasks, closing them, and creating sub-agents.\n - Sanitization: The documentation in
SKILL.mddoes not describe any sanitization or validation logic for external inputs.
Audit Metadata