test-campaign

Pass

Audited by Gen Agent Trust Hub on Nov 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses shell commands to interact with the environment, specifically invoking testgen, git apply, npm exec, and dotnet for test generation and execution.
  • [DYNAMIC_EXECUTION]: The skill's primary function involves generating source code patches via the testgen tool and applying them to the filesystem to be executed by test harnesses. This pattern of code generation and execution is a common vector for injecting malicious logic if the generator is compromised or misinformed.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes repository source code and user-provided JSON files to inform its test generation logic, which introduces a vulnerability to instructions embedded in that data.
  • Ingestion points: Source code from the target repository, project configuration files, and JSON input files for the testgen tool.
  • Boundary markers: The instructions do not define specific delimiters or security headers to prevent the agent from following instructions found within the processed repository files.
  • Capability inventory: Filesystem modification via git apply and shell command execution using local development tools (npm, dotnet, testgen).
  • Sanitization: The skill relies on manual review steps, instructing the user to "Review the property reference... before generation" and "Review the proposed patch before applying it."
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 21, 2026, 03:03 AM