prompt-lookup
Warn
Audited by Gen Agent Trust Hub on Feb 15, 2026
Risk Level: MEDIUMPROMPT_INJECTION
Full Analysis
- Indirect Prompt Injection (MEDIUM): The skill's primary function is to fetch and process prompt templates from an external source (
prompts.chat), which constitutes an attack surface for indirect instructions. - Ingestion Points: Untrusted data enters the agent context through the
search_promptsandget_prompttools inSKILL.md. - Boundary Markers: Absent. The skill does not instruct the agent to wrap retrieved content in delimiters or ignore instructions embedded within the fetched templates.
- Capability Inventory: The skill focuses on retrieval and enhancement (
improve_prompt). While it does not have direct file-write or shell execution capabilities, the output (prompts) is specifically designed to guide the agent's future logic, creating a high-influence surface. - Sanitization: Absent. There is no mention of validating or filtering the content retrieved from the external API.
Audit Metadata