newsletter-signal-scanner
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPERSISTENCE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external newsletter emails. This creates a surface for indirect prompt injection where malicious instructions embedded in a newsletter could potentially influence the agent during signal extraction or digest generation. * Ingestion points: AgentMail inbox (SKILL.md, Phase 1). * Boundary markers: None present. * Capability inventory: File system writes to the intelligence folder. * Sanitization: Only HTML-to-plain-text conversion is performed.
- [PERSISTENCE]: The skill provides documentation for scheduling recurring execution via crontab (SKILL.md, Scheduling section), which is a standard method for maintaining persistence in automation tasks.
Audit Metadata