The Agent Skills Directory

[SAFE]: No malicious patterns detected. The skill consists of instructional markdown and standard integration code examples targeting official Apideck and BambooHR services.
[INDIRECT_PROMPT_INJECTION]: The skill documents methods for reading untrusted data from BambooHR (e.g., employee records and custom fields).
Ingestion points: Data enters the agent via apideck.hris.employees.list and related methods in SKILL.md.
Boundary markers: None explicitly defined in the provided code snippets.
Capability inventory: No active code execution, file system writes, or dangerous capabilities are included in the skill body.
Sanitization: Standard API responses are processed; however, the skill itself does not implement additional sanitization beyond what is provided by the agent platform.
[DATA_EXPOSURE_AND_EXFILTRATION]: While the skill mentions access to sensitive fields like SSN and DOB, it explicitly notes that these require elevated admin permissions and uses environment variables (process.env.APIDECK_API_KEY) for credential management, which is a safe practice.

bamboohr