The Agent Skills Directory

[CREDENTIALS_UNSAFE]: Hardcoded credentials admin@test.com / SecurePass1234 are documented within the prerequisites section of the skill.
[COMMAND_EXECUTION]: The skill instructs the agent to execute several shell commands to manage the development environment and run tests:
npm run kill:e2e-ports to terminate existing processes.
npm run build and npm run start:umbraco to prepare and launch the application.
A complex health check loop using curl and seq: for i in $(seq 1 24); do if curl -sk -o /dev/null -w '%{http_code}' https://localhost:44374 --max-time 5 | grep -q '200\|302'; then ... fi; done.
npm run test:e2e and npx playwright test for test execution.
[INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted local data to influence agent behavior:
Ingestion points: The agent reads contents of demo-site/, .csproj files, Playwright test results, and standard error outputs from failed commands (e.g., in Step 6 to "diagnose" failures).
Boundary markers: No delimiters or safety instructions are used to distinguish between data being analyzed and the agent's instructions.
Capability inventory: The skill possesses significant capabilities including network operations via curl and arbitrary command execution via npm/npx.
Sanitization: There is no evidence of sanitization or content validation for the files or tool outputs processed during the diagnosis phase.

run-e2e