bump-openai-version
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- Standard Command Execution: The skill involves executing build and test commands, such as
dotnet restore,dotnet build, anddotnet test. These are standard operations for its intended purpose of upgrading and validating package versions within a development environment. - Network Probing of Trusted Sources: The instructions use
curlto verify package availability on the official .NET engineering Azure Artifacts feed (dnceng). This is a common practice for validating dependency availability before modification and targets a well-known service associated with the repository infrastructure. - Indirect Prompt Injection Surface: The workflow requires the agent to read external data from GitHub Pull Request descriptions and NuGet package metadata. While this introduces a potential surface for indirect instructions, the skill includes checkpoints requiring user approval before acting on findings and provides clear boundaries for the agent's actions.
- Credential Protection Measures: The instructions explicitly forbid the agent from requesting, printing, or committing credentials. This proactive instruction mitigates risks associated with sensitive data exposure during integration testing.
Audit Metadata