skills/yogsoth-ai/de-anthropocentric-research-engine/classify-falsification-verdict/Gen Agent Trust Hub
classify-falsification-verdict
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is composed strictly of descriptive text and logical instructions for an AI agent. It does not include any scripts, external dependencies, or commands that interact with the host system or network.
- [INDIRECT_PROMPT_INJECTION]: Ingestion points: Untrusted data enters the agent context via the 'Input contract' fields (claim, falsifier, test_result, adequacy_assessment) defined in SKILL.md. Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded commands. Capability inventory: No exploitable capabilities (such as subprocess calls, file writing, or network operations) are present across the skill. Sanitization: No sanitization or filtering logic is present, though the 'Quality gates' provide a logical structure for processing inputs.
Audit Metadata