dotnet-maui-doctor
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill instructs the agent to "Run through ALL tasks autonomously" and "Iterate until complete or no further actions possible." This instruction encourages the agent to proceed through environment modifications without requesting individual user confirmation for each step, which reduces oversight.
- [COMMAND_EXECUTION]: The skill dynamically constructs and executes shell commands (such as
dotnet workload installandsdkmanager) using versions and package names parsed from external JSON files. This creates a dependency on the integrity of the remote data sources for safe command execution. - [INDIRECT_PROMPT_INJECTION]: The skill ingests configuration data from external URLs (Microsoft's release metadata and NuGet's manifest API) to determine its execution flow and command arguments. This ingestion of untrusted data represents an attack surface where a compromised or malformed response could be used to inject malicious parameters into the shell commands.
- [EXTERNAL_DOWNLOADS]: The skill fetches .NET release information from
dotnetcli.blob.core.windows.netand downloads workload manifests fromapi.nuget.org. These are well-known, official sources for .NET development tooling and are used to ensure environment currency.
Audit Metadata