The Agent Skills Directory

[COMMAND_EXECUTION]: The skill contains several Bash and Python scripts designed to automate interactions with AI models via the belt command-line interface. This is the intended functionality for building automated workflows.
[EXTERNAL_DOWNLOADS]: Includes references to installation documentation hosted on the vendor's official GitHub organization (inference-sh). These are trusted setup instructions for the required CLI tools.
[PROMPT_INJECTION]: The provided workflow templates ingest external data from files and command-line inputs into AI model prompts. This represents a potential surface for indirect prompt injection if the inputs originate from untrusted sources.
Ingestion points: Uses content from local text files in data_processing.sh and input variables in conditional_workflow.sh.
Boundary markers: No explicit delimiters or instructions are provided in the scripts to isolate untrusted content from the system instructions.
Capability inventory: Access to the local file system, network capability via curl for alerting, and model execution via the belt CLI.
Sanitization: The Python script uses json.dumps for input formatting, while the Bash scripts use direct variable interpolation within JSON strings.

ai-automation-workflows