automate-e2e-tests
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses
rgandfindto discover project structure and search for existing test specifications. It also executes various standard testing frameworks including Playwright, Cypress, WebdriverIO, Appium, and Maestro. These tools are used according to established project documentation and within their intended functional scope.- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests untrusted data to drive test generation and execution. - Ingestion points: The agent is instructed to read external requirements, tickets, and implementation diffs from the project environment.
- Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between valid task data and potentially malicious embedded commands within these inputs.
- Capability inventory: The agent has the capability to write new test files and execute shell commands via various testing runners (e.g.,
npx,gradlew,xcodebuild). - Sanitization: The instructions do not specify any validation, escaping, or filtering of the input data before it is interpolated into prompts or used to generate code.- [PROMPT_INJECTION]: The 'Self-Improvement' section contains instructions for the agent to 'patch this skill or the configured project-local version' based on its performance. This allows for modification of the agent's instructions, which is an agent-behavior override pattern.
Audit Metadata