e2e-debugger

Pass

Audited by Gen Agent Trust Hub on Nov 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill performs local command execution to run test suites and analysis scripts. It uses npx for Playwright and Cypress tests and executes a specific shell script via /bin/bash -p <e2e-reviewer dir>/scripts/scan.sh. These commands are scoped to the local environment but allow the agent to interact with the system's shell.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources that are not fully trusted, such as test specification files, application source diffs, and logs from test failures. This information is used to drive the agent's logic for repairing code and generating new test commands.
  • Ingestion points: Specification files, test runner output, and application diffs read from the project directory.
  • Boundary markers: The instructions do not employ specific delimiters or instruction-ignore blocks to separate untrusted data from the agent's operational logic.
  • Capability inventory: The skill can execute shell commands and write to various test-related files in the repository.
  • Sanitization: There is no evidence of sanitization or validation of the ingested content, although the skill explicitly warns that page content is untrusted when using sub-agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 19, 2026, 08:22 AM