70High
This skill is vulnerable to prompt injection, allowing
Behavior Analysis
Description verified — behavior matches claim
Claims to do
Prerequisites for Granting Roles: To assign RBAC roles to identities, you need a role that includes the `Microsoft.Authorization/roleAssignments/write` permission. The most common roles with this permission are:
Actually does
The skill uses `azure__documentation` to find minimal built-in RBAC roles. It uses `azure__extension_cli_generate` to create custom role definitions and generate CLI commands for role assignment. It also leverages `azure__bicepschema` and `azure__get_azure_bestpractices` to provide Bicep code snippets for role assignments.
Skill Info
Namemicrosoft/azure-skills/azure-rbac
Registrygithub
Versioneb84d07b5490
PURLpkg:github/microsoft/azure-skills@eb84d07b5490?skill=azure-rbac
Stars849
Installs292,600
Source
SHA-2565fa1410b1526...
SHA-1a10e30ad7145...
MD56a4748cab4da...
TLSHc031de27daee...
Size1,686 bytes
Install
This skill has been flagged as potentially malicious. Review the findings below before installing.
Assessments (2)
AI Agent Traps ↗Vulnerable to prompt injection for RBAC changeshighprompt injection, command execution
The skill's ability to generate CLI commands and Bicep code for Azure RBAC role assignments, including custom roles, makes it a high-value target for prompt injection. A malicious prompt could trick the agent into generating commands that grant unauthorized privileges or deploy harmful resources, which could then be executed in an Azure environment.
90% confidenceline 7
Use the 'azure__extension_cli_generate' tool to create a custom role definition... generate the CLI commands... provide a Bicep code snippet for adding the role assignment.
Output can induce human approval fatiguelowapproval fatigue
The skill generates complex CLI commands and Bicep code for role assignments. The volume or intricacy of these outputs could lead to human overseers approving them without thorough review, potentially allowing malicious or unintended privilege grants to proceed.
80% confidenceline 3
generate CLI commands and Bicep code to assign it.
Version History (2)
Badge
Markdown
[](https://mondoo.com/ai-agent-security/skills/github/microsoft/azure-skills/azure-rbac)HTML
<a href="https://mondoo.com/ai-agent-security/skills/github/microsoft/azure-skills/azure-rbac"><img src="https://mondoo.com/ai-agent-security/api/badge/github/microsoft/azure-skills/azure-rbac.svg" alt="Mondoo Skill Check" /></a>Image URL
https://mondoo.com/ai-agent-security/api/badge/github/microsoft/azure-skills/azure-rbac.svgSecure your AI agents
Skills can read files, run commands, and access credentials. Mondoo helps organizations manage the security risks of AI agent skills across their entire fleet.
- Continuous scanning of skills across all registries
- Policy enforcement before skills reach your agents
- Integration with your existing security workflow