The Agent Skills Directory

[SAFE]: The skill provides utility scripts (management_review_tracker.py and quality_effectiveness_monitor.py) that process local data files (JSON/CSV) to generate quality reports. These scripts use only Python standard libraries and do not perform network requests, file system modifications (beyond reading specified inputs), or dynamic code execution.
[SAFE]: Documentation and templates are consistent with ISO 13485:2016 standards and do not contain hidden instructions, prompt injections, or obfuscated content.
[INDIRECT_PROMPT_INJECTION]: The scripts ingest external data from CSV and JSON files provided by the user. While this creates an entry point for data that the AI agent will process, the scripts themselves do not possess high-privilege capabilities like shell execution or network access, making the risk of exploitation negligible.
Ingestion points: management_review_tracker.py (via --data JSON) and quality_effectiveness_monitor.py (via --metrics CSV).
Boundary markers: None explicitly present in the data ingestion logic, though the scripts strictly parse structured formats (JSON/CSV).
Capability inventory: File read, stdout printing. No subprocess, exec/eval, or network capabilities detected.
Sanitization: Standard json.load() and csv.DictReader() are used.

quality-manager-qmr