memory-literary-analysis
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill outlines a structured workflow for literary analysis using a set of dedicated memory management tools (such as
create_memory_project,write_note, andschema_validate). These tools are used within their intended scope to organize and validate textual data. References to external sources like Project Gutenberg for acquiring literary texts are documented neutrally as a standard source for public domain works. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external literary texts to generate structured notes, creating a potential surface for indirect prompt injection if a source text contains malicious instructions aimed at the processing agent.
- Ingestion points: External literary source texts processed during the chapter-by-chapter workflow (SKILL.md).
- Boundary markers: The skill employs structured schemas and observation categories (e.g.,
[summary],[event]) which provide a framework for data extraction, though explicit string sanitization for input text is not mentioned. - Capability inventory: Tools for project creation, note writing/editing, and schema validation (
write_note,edit_note,schema_validate,create_memory_project). - Sanitization: Not explicitly detailed in the instructional workflow for processing raw text.
Audit Metadata