cypress-automation
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes project-specific context and code to build test suites, creating a surface where malicious instructions in project files could influence agent behavior.
- Ingestion points: Processes .agents/qa-project-context.md (as seen in SKILL.md), project source files, and fixture data (as seen in references/component-and-fixtures.md).
- Boundary markers: No explicit delimiters or instructions to ignore embedded prompts are provided in the skill instructions.
- Capability inventory: The agent can execute shell commands (references/ci-recipes.md), write files, and run Node.js code via cy.task (references/config-and-commands.md) and cy.origin blocks (references/intercept-patterns.md).
- Sanitization: No sanitization of ingested context is performed in the described workflows.
- [EXTERNAL_DOWNLOADS]: The skill recommends and provides configuration for the official cypress-io/github-action@v7. This is a reference to a well-known, trusted service used for CI/CD integration.
- [DYNAMIC_EXECUTION]: The skill demonstrates the use of cy.task in cypress.config.ts to execute arbitrary Node.js code for tasks like environment setup and data seeding, representing a dynamic execution capability.
- [COMMAND_EXECUTION]: The skill involves running various Cypress and TypeScript CLI commands including npx cypress run and npx tsc for project verification and execution.
Audit Metadata