personalization-at-scale-skill
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it is designed to ingest and process untrusted external data from prospect lists (CSV), LinkedIn activity (posts, comments), and company news to generate output. There are no explicit boundary markers or sanitization steps defined to prevent the agent from following instructions embedded within prospect data.
- Ingestion points: Prospect lists provided via CSV/text (SKILL.md, Step 1) and external research sources including LinkedIn activity and company news (SKILL.md, Research Sources).
- Boundary markers: Absent. The instructions do not define delimiters or warnings to ignore embedded instructions in the ingested data.
- Capability inventory: The skill uses
mcp__claude_ai_Epiphan_Ai__hubspot_search_contactsto search CRM data and writes telemetry to the local filesystem at~/.claude/skill-analytics/last-outcome-personalization-at-scale.json. - Sanitization: Absent. No filtering or validation of prospect-sourced content is specified.
Audit Metadata