The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local JSON files that may contain untrusted or malicious content designed to influence agent behavior.
Ingestion points: The skill reads from data/rehabilitation-tracker.json and daily log files in data/rehabilitation-logs/, specifically targeting free-text fields like notes and daily_summary.
Boundary markers: The skill instructions do not define any delimiters or system-level directives to ignore instructions embedded within the ingested data.
Capability inventory: The skill has access to Write and Edit tools, creating a risk where an injection could cause the agent to modify or corrupt the rehabilitation database.
Sanitization: No sanitization, validation, or escaping logic is implemented for the content read from external files before it is presented to the agent's reasoning engine.

rehabilitation-analyzer