csharp-tdd-develop
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- PROMPT_INJECTION (LOW): The skill interpolates user-provided
$ARGUMENTSdirectly into the instructions for a sub-agent. A user could provide a crafted description that overrides the sub-agent's logic or forces it to perform unintended file operations. - PROMPT_INJECTION (LOW): Indirect prompt injection surface via project metadata.
- Ingestion points: The
scripts/test-detector.jsscript reads and extracts package names from.csprojfiles within the user's workspace. - Boundary markers: Absent. The output of the detector script is presented directly to the orchestrator agent without sanitization or delimiters.
- Capability inventory: The agent has
Bash,Write,Edit, andTaskcapabilities, which are high-impact if manipulated. - Sanitization: None. The skill assumes the contents of
.csprojfiles are legitimate XML/metadata. - COMMAND_EXECUTION (LOW): The skill uses
Bash(dotnet test *). While essential for the primary purpose (TDD), executing tests on untrusted projects can lead to arbitrary code execution if the test suites contain malicious setup code or life-cycle hooks.
Audit Metadata