schema-author
Pass
Audited by Gen Agent Trust Hub on Aug 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a comprehensive suite of
gbrain schemaCLI tools and standard git commands (git add,git commit,git push) to manage system configuration and version control. - [DATA_EXFILTRATION]: The workflow involves a
git pushoperation which transmits local schema configurations to a user-configured remote repository. This is a standard and expected functionality for a development tool. - [INDIRECT_PROMPT_INJECTION]: The skill incorporates an analysis phase where it ingests the user's existing page corpus to suggest new schema types, representing an inherent surface for indirect prompt injection.
- Ingestion points: Content is read from the user's local page corpus during the
gbrain schema detectandgbrain schema suggestoperations (SKILL.md). - Boundary markers: The instructions do not specify the use of delimiters or ignore-instructions markers when the agent processes the corpus data.
- Capability inventory: The skill can execute shell commands, write to local configuration files in
~/.gbrain/, and perform network operations via git. - Sanitization: There are no documented steps for sanitizing or filtering the content ingested from the corpus before analysis.
Audit Metadata