The Agent Skills Directory

[SAFE]: The skill operates entirely within its stated scope, utilizing local file access for record-keeping. No signs of data exfiltration, remote code execution, or persistence mechanisms were detected.
[PROMPT_INJECTION]: The skill processes untrusted natural language input from users to populate medical records. This represents an indirect prompt injection surface. However, the risk is mitigated by the limited capabilities of the skill, which is restricted to local file operations.
Ingestion points: Natural language symptom descriptions provided in the add operation.
Boundary markers: Absent; the skill does not use specific delimiters to isolate user input from its internal logic.
Capability inventory: Read and Write tool access used for managing JSON files in the data/ directory.
Sanitization: Not documented; the skill converts colloquialisms to medical terms but does not explicitly sanitize for malicious instructions.

symptom