causality-engine
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides a structured framework for analyzing plot development and narrative logic. It is purely instructional and does not contain any executable scripts, remote downloads, or credential-accessing patterns.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted narrative content provided by users.
- Ingestion points: The skill ingests user-provided plot points, character actions, and story summaries for causal analysis.
- Boundary markers: There are no explicit instructions to the agent to treat narrative data as distinct from instructions (e.g., using delimiters or specific 'ignore instructions' wrappers).
- Capability inventory: The skill's primary side effect is writing analysis results to a local file (
causality/CAUSAL_CHAIN.md) and coordinating with other narrative engines. It possesses no capabilities for network exfiltration, arbitrary command execution, or system-level modifications. - Sanitization: No content sanitization is defined for the input narrative data.
Audit Metadata