The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a local Python script (scripts/prereg_check.py) to perform structured audits of research plans. The script is restricted to the Python standard library, performs no network operations, and does not invoke sub-processes or dynamic code execution.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data via JSON input files (plan.json and actual.json).
Ingestion points: External data enters the agent context through the output of the prereg_check.py script, which reads user-supplied JSON files via command-line arguments.
Boundary markers: The skill does not explicitly define prompt boundary markers for the data ingested from the script output, relying instead on the structured nature of the JSON-to-text conversion.
Capability inventory: The skill possesses Read, Write, Edit, and Bash(python:*) capabilities. The Python script performs local file reading and basic JSON parsing.
Sanitization: The Python script uses json.load(), which provides inherent safety against code injection during data ingestion compared to unsafe deserialization methods like pickle or eval.

alterlab-preregistration-discipline