Reddi Self Improving Agent

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:15 PM UTC
• Version:
  1.0.13

Overview

This skill is a disclosed local learning and reminder helper; its main risk is that users may persist sensitive context or enable broad optional hooks.

Install only if you want agents to keep local learning notes and future-session reminders. Avoid logging secrets, tokens, private transcripts, customer data, or sensitive stack traces; review entries before promoting them into shared agent guidance files, and prefer project-level or matcher-scoped hooks over global always-on hooks.

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 (1)

Vague Triggers

Medium

• Confidence: 89% confidence
• Finding: The user-level configuration installs the hook globally with an empty matcher, causing automatic execution across all prompts and projects. Because hook scripts run with the same permissions as the agent environment, this broad persistent trigger increases the blast radius of any future script flaw, prompt-sensitive behavior, or unintended data capture across unrelated repositories.

VirusTotal

65/65 vendors flagged this skill as clean.