skills/eirkkk/winland-android/report/Gen Agent Trust Hub

report

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local shell commands including npx playwright, ls, and tee to manage test execution and check for result artifacts.\n- [DATA_EXFILTRATION]: Utilizes curl to transmit test summaries to an external Slack webhook. This destination is dynamically configured via the SLACK_WEBHOOK_URL environment variable, which is a safe practice for user-defined integrations.\n- [PROMPT_INJECTION]: The skill ingests test output (logs and JSON) which may contain untrusted data such as error messages or test names, creating a surface for indirect prompt injection.\n
  • Ingestion points: Test output logs and Playwright JSON reports from test-results/.\n
  • Boundary markers: No explicit delimiters or boundary markers are used to wrap the interpolated test results in the markdown template.\n
  • Capability inventory: Command execution (npx), network access (curl), and file system operations.\n
  • Sanitization: There is no evidence of sanitization or escaping of test output content before it is included in reports or sent to external webhooks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 08:33 AM
Security Audit — agent-trust-hub — report