dialogue-authoring
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: No safety bypasses or instructions to ignore previous rules were found. The skill focuses strictly on technical game modding tasks.
- [DATA_EXFILTRATION]: No network access or sensitive file reading patterns were identified. The skill operates on local game plugin data.
- [DYNAMIC_EXECUTION]: The skill manages the compilation of Papyrus scripts using local tools. This is a core function of the skill and does not involve untrusted remote execution.
- [INDIRECT_PROMPT_INJECTION]: The skill audits and interprets external game records.
- Ingestion points: Dialogue records are retrieved using
housecarl_read_recordandhousecarl_cross_plugin_queryin SKILL.md. - Boundary markers: Not specifically defined in the instructions.
- Capability inventory: Includes record creation, field modification, and script compilation in SKILL.md.
- Sanitization: Not explicitly required for this domain-specific task.
- [OBFUSCATION]: No obfuscated URLs, base64-encoded commands, or hidden characters were found.
Audit Metadata