specialist-specialized-skill-builder
Pass
Audited by Gen Agent Trust Hub on Aug 15, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill takes user input and interpolates it directly into a prompt for a sub-agent that has filesystem write access.\n
- Ingestion points: The variable
{the user's question}inSKILL.mdacts as the entry point for untrusted data.\n - Boundary markers: There are no explicit delimiters (such as XML tags or triple quotes) used to isolate the user's input from the sub-agent's instructions, increasing the risk of command confusion.\n
- Capability inventory: The spawned sub-agent is authorized to write
SKILL.mdfiles and updateplugin.jsonconfigurations in the filesystem.\n - Sanitization: No sanitization, filtering, or validation logic is present to handle the user-provided string before it influences the sub-agent's execution context.
Audit Metadata