The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes shell scripts scripts/test-beat.sh and scripts/validate-schedule.sh that automate the testing and validation of the Celery environment by executing local commands and dynamically generating Python code to inspect application configuration.
[PROMPT_INJECTION]: The template templates/dynamic-schedules.py implements a pattern for loading task schedules from external JSON files, which introduces an indirect prompt injection surface where a malicious configuration file could be used to execute unauthorized tasks within the task queue.
Ingestion points: External JSON configuration files (e.g., schedules.json) loaded via load_schedule_config.
Boundary markers: None identified in the provided templates to isolate external configuration from the execution logic.
Capability inventory: The skill uses sender.add_periodic_task and app.signature which are capable of triggering any registered task.
Sanitization: No validation or sanitization logic is implemented for the task signatures or arguments ingested from the external files.

beat-scheduling