The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data fetched from an external SailPoint IdentityIQ instance, which constitutes an indirect prompt injection surface.
Ingestion points: The scripts/agent.py file fetches user, entitlement, and certification data from the SailPoint SCIM and REST APIs.
Boundary markers: No explicit boundary markers or "ignore instructions" delimiters are used when processing the data for reporting.
Capability inventory: The script has network access to the SailPoint API and outputs data to the agent's console, which can then be interpreted by the LLM.
Sanitization: The script performs no sanitization or filtering of the content returned by the external API before presenting it to the agent.
[COMMAND_EXECUTION]: The provided Python script (scripts/agent.py) accepts a username and password directly as command-line arguments. This is a suboptimal security practice as process arguments are often visible to other users or logged by the operating system, potentially exposing credentials.

performing-entitlement-review-with-sailpoint-iiq