The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes test result files (JSON format) generated by the Playwright testing framework, creating a surface for indirect instructions to influence agent behavior.
Ingestion points: Test result JSON output from npx playwright test and historical data stored in the test-reports/ directory.
Boundary markers: Absent; test failure details and error messages are interpolated directly into markdown and curl payloads using template placeholders (e.g., {{error}}).
Capability inventory: The skill can execute shell commands (curl, npx), perform filesystem operations (reading results, writing reports), and send data over the network.
Sanitization: No sanitization or validation of the test result content is performed before generating reports or triggering webhooks.
[DATA_EXFILTRATION]: The skill checks for and utilizes sensitive environment variables (TESTRAIL_URL, SLACK_WEBHOOK_URL) to send test metadata to external endpoints.
It uses curl to POST test summary data to a dynamic domain provided in the Slack webhook variable.
It invokes an external tool integration (/pw:testrail push) that utilizes the TESTRAIL_URL configuration.
[COMMAND_EXECUTION]: The skill relies on shell command execution for its core functionality.
Executes npx playwright test to run test suites and generate report data.
Uses curl for network-based reporting.
Employs common utilities like tee, ls, and echo for logging and environment discovery.

report