moai-workflow-templates
Warn
Audited by Gen Agent Trust Hub on Feb 18, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [Dynamic Execution] (MEDIUM): The
template-optimizer.mdmodule performs automated project restoration that modifies local files. - Evidence: Functions
restore_from_backupandsmart_mergeinmodules/template-optimizer.mdscan the.moai-backups/directory and apply stored content to the active project. - Risk: This mechanism could be exploited to persist malicious code or configuration if the backup directory is compromised.
- [Indirect Prompt Injection] (LOW): The feedback mechanism in
modules/feedback-templates.mdprocesses untrusted data. - Ingestion points: User input from the
/moai:9-feedbackcommand (modules/feedback-templates.md). - Boundary markers: Absent; user input is interpolated directly into templates without delimiters.
- Capability inventory: Automatic GitHub issue creation (Network Write).
- Sanitization: Absent; no validation or escaping shown for user-provided descriptions.
- [Data Exposure & Exfiltration] (LOW): Automated network communication with GitHub.
- Evidence:
modules/feedback-templates.mdtriggers automated GitHub issue creation based on user-supplied parameters.
Audit Metadata