The Agent Skills Directory

[SAFE]: No malicious patterns, prompt injections, or unauthorized data exfiltration attempts were detected. The skill's behavior aligns with its stated purpose of providing HR and organizational leadership advice.
[INDIRECT_PROMPT_INJECTION]: The skill includes an attack surface for indirect prompt injection by ingesting external data, though no active exploitation was found.
Ingestion points: Reads from ~/.claude/company-context.md and user-supplied JSON files (via --input flags in scripts).
Boundary markers: Absent; the skill does not define specific delimiters for external context.
Capability inventory: Includes three Python scripts that perform local data processing and reporting; no network or high-privilege operations are requested.
Sanitization: Not present; the skill relies on the underlying model's safety filters for processing ingested text.

chro-advisor