skills/garethmanning/education-agent-skills/ai-hallucination-fact-check-protocol/Gen Agent Trust Hub
ai-hallucination-fact-check-protocol
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input by interpolating it directly into the prompt template, creating a vulnerability surface where malicious instructions in the input could attempt to override the skill's behavior.
- Ingestion points: The
ai_output_contextandstudent_levelparameters inSKILL.mdare used to populate the generation prompt. - Boundary markers: Absent. The prompt template does not use delimiters (such as triple backticks or XML tags) or specific instructions to the model to ignore embedded commands within the user-provided data.
- Capability inventory: None. The skill performs text generation only and does not possess capabilities for file writing, network access, or command execution.
- Sanitization: Absent. No input validation or filtering logic is present to identify or neutralize potential injection attempts within the user data.
Audit Metadata