build-backend

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION] (SAFE): The skill invokes the dotnet CLI to run a local script (.github/skills/build-backend/build-backend.cs). This operation is restricted to the repository's own code and is the intended primary purpose of the skill.
  • [INDIRECT_PROMPT_INJECTION] (SAFE): The skill processes output from the build process. 1. Ingestion points: Build error messages in stderr/stdout. 2. Boundary markers: Absent. 3. Capability inventory: Command execution (dotnet). 4. Sanitization: The skill captures and reports errors but does not pass them to high-risk sinks. Since the data originates from local source code compilation, the risk of adversarial injection is negligible.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 06:24 PM