Copy My Profile

• Outcome: Pass
• Latest audit: Jun 20, 2026 at 6:09 PM UTC
• Version:
  1.0.0

Overview

This skill is a local profile-export helper that handles sensitive memory data, but its behavior is disclosed, purpose-aligned, and not hidden or destructive.

Install only if you are comfortable letting the agent read your AI memory/config files and create a consolidated profile. Review and redact my-profile.md before saving it in synced folders, committing it, or importing it into ChatGPT, Claude, Cursor, Codex, or another tool's long-term memory.

SkillSpector

By NVIDIA

Vulnerability Patterns

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

Findings (8)

Vague Triggers

Medium

• Confidence: 92% confidence
• Finding: The trigger phrase 'Generate my profile' is generic enough to plausibly appear in normal conversation, which increases the chance of accidental invocation. In the context of this skill, accidental activation is more dangerous because the described behavior includes reading long-term memory files and producing a consolidated profile containing personal preferences and identity/profession data.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The README does mention privacy in general terms, but it does not provide a prominent, explicit warning near usage that the skill reads memory files from multiple agent tools and writes a portable profile file containing personal preference and identity/profession data. That omission can lead users to invoke the skill without understanding the scope of local data access and the sensitivity of the generated artifact, increasing the risk of unintended disclosure or mishandling.

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The skill instructs the agent to write `./my-profile.md` without an explicit confirmation step or warning that it will modify the filesystem. Even though the write target is predictable, creating files containing extracted profile data can expose sensitive information locally, overwrite user expectations, or leave artifacts in repositories or synced folders.

Missing User Warnings

Medium

• Confidence: 95% confidence
• Finding: The file instructs users to copy profile data into long-term memory stores and local configuration files across multiple tools, but it does not warn that the exported profile may contain sensitive personal, professional, or project information that will persist beyond the current session. That omission can lead users to disclose data they would not otherwise store durably, increasing privacy and data-retention risk across local and vendor-managed systems.

Missing User Warnings

Medium

• Confidence: 97% confidence
• Finding: The ChatGPT import instructions tell users to ask the service to remember profile information for future conversations, but they do not clearly warn that this may place personal data into vendor-managed memory outside the user's local environment. Without an explicit caution to review and minimize sensitive content first, users may unintentionally persist private information in a third-party retention system.

Ssd 3

Medium

• Confidence: 95% confidence
• Finding: The core skill behavior is to aggregate long-term memory from multiple files into a portable profile for use in other tools, which creates a concentrated export of potentially sensitive personal instructions, identity details, work context, and project history. This significantly increases exposure risk because data that was previously scattered across tool-specific memory stores becomes normalized, copied, and easier to exfiltrate or paste into less trusted environments.

Ssd 3

Medium

• Confidence: 94% confidence
• Finding: Reading recent daily memory logs to 'capture recent context' expands collection beyond stable preference files into potentially sensitive, time-specific notes, conversations, and ephemeral personal or workplace data. This broadens the chance of extracting secrets, confidential project details, or private behavioral information that the user did not intend to export into a reusable profile.

Ssd 3

Medium

• Confidence: 96% confidence
• Finding: The skill directs both writing the compiled profile to disk and returning it inline, creating two disclosure channels for the same sensitive aggregation. Duplicating output increases the chance of accidental exposure through terminal history, chat transcripts, synced folders, version control, or downstream copy/paste into other tools.

VirusTotal

63/63 vendors flagged this skill as clean.