writing

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill utilizes a set of SurrealDB query templates in pre-write-query.surql to facilitate context gathering for writers. These queries are scoped to creative and technical writing entities, such as characters, locations, events, and manuscript sections, and do not interact with sensitive system files or credentials.
  • [SAFE]: The supplemental files citation.md, dialogue.md, and voice.md contain purely instructional content designed to improve writing quality. They do not contain executable code, obfuscated payloads, or network exfiltration patterns.
  • [SAFE]: No evidence of prompt injection or attempts to override agent safety protocols was detected. The workflow is designed for author assistance and adheres to creative continuity principles without requesting elevated privileges or persistent access.
  • [SAFE]: The data ingestion surface (retrieving content from the database for prose generation) is a standard design for database-backed writing tools and, in this context, processes user-provided manuscript data without risk of external exploitation.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 08:06 AM