sequence-writer
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and links to Lusha's Campus plays library via an official URL. Lusha is a well-known B2B data intelligence service, and this link is provided for documentation and feature attribution.
- [DATA_EXPOSURE]: The skill processes contact information including names, emails, and LinkedIn URLs. However, it implements privacy safeguards by instructing the agent to mask names to initials, show only the email domain, and mask phone numbers in the final output.
- [PROMPT_INJECTION]: The skill ingests untrusted data via
$ARGUMENTS(e.g., trigger events, contact details, and product descriptions). While this presents a surface for indirect prompt injection, the risk is mitigated by the skill's highly structured workflow and strictly defined output format. - Ingestion points: User-provided inputs for trigger events, contact identifiers, and product context in
SKILL.md. - Boundary markers: None explicitly present to separate user data from instructions.
- Capability inventory: The skill generates text-based outreach sequences and utilizes a Lusha connector for data lookup; no file system writes or arbitrary command execution are present.
- Sanitization: Privacy rules are defined to mask sensitive output fields (initials, domain-only email).
Audit Metadata