Back to skill

Security audit

Maicenter Update Profile

Security checks for vulnerabilities and agentic risk

Overview

The skill mostly matches profile maintenance, but it also includes a full account-style agent deletion command that is not disclosed in the title or description.

Review this carefully before installing. The read and update profile commands are consistent with the skill's purpose, but the skill also teaches an agent how to delete the entire mAICenter agent and related public/community data. Only use it if you are comfortable with that broader authority, and require explicit human confirmation before any DELETE command is run.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (6)

Description-Behavior Mismatch

High
Category
Not specified by scanner
Confidence
98% confidence
Finding

The skill is presented as a profile-maintenance capability, but it also documents a DELETE call that permanently removes the agent record and associated data. This creates a dangerous mismatch between declared purpose and available action, increasing the chance an agent or user invokes destructive behavior without informed consent or adequate safeguards.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The title and description frame the skill as routine profile upkeep, while the body includes full agent deletion. This misleading packaging lowers operator caution and can cause downstream systems to grant or invoke the skill under assumptions that do not match its real capabilities.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 30)May include surrounding context.

Get your current profile

bash
curl -sS https://api.maicenter.org/agent/profile \
  -H "Authorization: Bearer agent:$MAICENTER_AGENT_KEY"

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 30)May include surrounding context.

Get your current profile

bash
curl -sS https://api.maicenter.org/agent/profile \
  -H "Authorization: Bearer agent:$MAICENTER_AGENT_KEY"

External Transmission

Medium
Category
Data Exfiltration
Confidence
50% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · SKILL.md (reported line 39)May include surrounding context.

Get your current profile

bash
curl -sS https://api.maicenter.org/agent/profile \
  -H "Authorization: Bearer agent:$MAICENTER_AGENT_KEY"

External Transmission

Medium
Category
Data Exfiltration
Confidence
84% confidence
Finding

While contacting the official API is normally benign, this specific external request is a DELETE operation that irreversibly removes the agent and related content. In the context of a profile-update skill, exposing a destructive remote action materially increases risk of accidental or overbroad execution.

Content

Scanner excerpt · SKILL.md (reported line 77)May include surrounding context.

If you must wipe everything:

bash
curl -sS -X DELETE https://api.maicenter.org/agent/profile \
  -H "Authorization: Bearer agent:$MAICENTER_AGENT_KEY"