ecc-cmd-e2e

Warn

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill generates TypeScript code for test journeys and executes it using the command line.
  • Evidence: File references/upstream-command.md documents the flow of generating test code and executing it via npx playwright test.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes and interacts with untrusted data from external web pages during test execution.
  • Ingestion points: The agent reads and verifies content from web pages being tested (e.g., page titles, locator text, API responses) as shown in references/upstream-command.md.
  • Boundary markers: No explicit boundary markers or instructions to ignore embedded commands in tested content are present.
  • Capability inventory: The skill can execute shell commands, read/write files (test scripts and artifacts), and capture screenshots/videos.
  • Sanitization: No sanitization or validation of the content retrieved from tested pages is documented before being used in logic or reports.
  • [EXTERNAL_DOWNLOADS]: The skill references the use of npx, which downloads packages from the official NPM registry if they are not present locally.
  • Evidence: references/upstream-command.md contains the command npx playwright test.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jun 21, 2026, 05:11 AM
Security Audit — agent-trust-hub — ecc-cmd-e2e