The Agent Skills Directory

[SAFE]: The skill implements a structured checklist for production deployment verification without introducing significant security risks. It encourages security best practices such as verifying secrets, auth isolation, and data durability.
[COMMAND_EXECUTION]: The skill provides an inline bash script to perform a canary health check on the deployed application. This script uses standard safety flags (set -euo pipefail) and validates the target URL against a specific regular expression.
[EXTERNAL_DOWNLOADS]: The checklist uses curl to probe the /healthz endpoint of the published Replit application. The target domain is restricted via regex to the well-known replit.app service.
[INDIRECT_PROMPT_INJECTION]: The skill reads and processes project files using the Read and Grep tools, which constitutes a potential injection surface for untrusted code to influence agent behavior.
Ingestion points: Project configuration and source files are inspected using the Read and Grep tools to verify settings and locate credentials.
Boundary markers: The instructions do not define explicit delimiters or instructions for the agent to ignore embedded commands in the files it reads.
Capability inventory: The skill has access to the Bash tool to perform network connectivity probes.
Sanitization: The skill mitigates risks by validating the health check URL with a regular expression and instructing the agent to grep for call sites without printing or displaying secret values.

replit-prod-checklist