The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill contains several templates that ingest untrusted data and interpolate it directly into AI prompts.
Ingestion points: The script conditional_workflow.sh accepts user-supplied text via command-line arguments. The script data_processing.sh reads the contents of local text files using cat and embeds them into a prompt.
Boundary markers: Absent. The scripts perform simple string interpolation without delimiters or instructions to ignore embedded commands.
Capability inventory: The skill utilizes the belt CLI to execute AI models, performs file operations via shell redirection, and includes an example (monitored_workflow.sh) that uses curl for network requests.
Sanitization: Absent. Content is passed directly to the model as part of the prompt string.
[PERSISTENCE]: The skill provides explicit instructions for establishing persistence through system scheduling tools. It includes crontab templates to execute automation scripts at regular intervals (daily, weekly, and every 6 hours).
[EXTERNAL_DOWNLOADS]: The skill references external resources for installation, specifically linking to a CLI installation guide hosted at https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md.
[COMMAND_EXECUTION]: The skill relies heavily on shell scripting for workflow orchestration. It provides multiple Bash script patterns and a Python script that uses subprocess.run to execute system commands via the belt CLI.

ai-automation-workflows