The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its reliance on data ingested from external files.
Ingestion points: Data is read from various files including data-example/nutrition-logs/*.json, fitness-tracker.json, sleep-tracker.json, hypertension-tracker.json, and diabetes-tracker.json.
Boundary markers: Absent; there are no delimiters or explicit instructions telling the agent to treat the contents of these files as data only rather than instructions.
Capability inventory: The skill is granted Read, Grep, Glob, and Write tool permissions, which could allow an attacker to influence filesystem state if malicious instructions were successfully injected via the data logs.
Sanitization: No documentation or code suggests that input data is validated, filtered, or sanitized before being processed by the agent.

nutrition-analyzer