Eo Ability Rag

• Outcome: Review
• Latest audit: May 29, 2026 at 9:32 PM UTC
• Version:
  1.0.0

Overview

This RAG skill is coherent, but it encourages automatic cross-project knowledge storage without enough privacy, consent, deletion, or scoping detail.

Use this only for knowledge you intentionally want stored and reused across projects or sessions. Do not index secrets, customer data, regulated data, or confidential project material unless you have verified where the knowledge base lives, who can search it, how to delete entries, and whether sharing can be scoped or disabled.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (2)

Vague Triggers

Medium

• Confidence: 87% confidence
• Finding: The skill states that knowledge indexing may be 'automatically triggered by EO' without clearly defining scope, consent, or what content is eligible for ingestion. In a cross-project RAG system, this ambiguity can cause unintended indexing of sensitive project data, creating unauthorized retention and later retrieval across contexts.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The description promotes cross-project knowledge sharing and continuous learning but does not warn users that project content may be persisted and reused beyond the original context. This is dangerous because users may unknowingly submit confidential, regulated, or customer data into a shared knowledge base, leading to data leakage across projects or sessions.

VirusTotal

65/65 vendors flagged this skill as clean.