The Agent Skills Directory

Command Execution via Helper Scripts: The skill employs custom scripts (retry-az-create.sh and retry-az-create.ps1) to wrap Azure CLI operations. These scripts use shell evaluation techniques to provide idempotent 'show-or-create' logic with automatic retries for transient network errors. This is a standard automation pattern for cloud infrastructure management.
Sensitive File Exclusion: During the packaging phase, the skill's instructions explicitly exclude sensitive directories and files, such as .env, .git, and .venv, from the deployment ZIP. This practice helps prevent the accidental exfiltration of local secrets or unnecessary metadata to the cloud environment.
Input Sanitization for Resource Naming: The skill includes dedicated scripts (generate-app-name.sh and generate-app-name.ps1) that sanitize folder names into valid Azure resource identifiers. By removing non-alphanumeric characters and enforcing length constraints, these scripts provide a layer of validation for parameters used in subsequent shell commands.
Framework Detection and Configuration: The skill analyzes project metadata files (like requirements.txt) to advisory detect frameworks (Flask, Django, FastAPI). This allows for the automatic configuration of appropriate startup commands, reducing manual configuration errors that could lead to service unavailability.

python-appservice-deploy