access-review
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external identity providers which could potentially be manipulated by an attacker to include instructions targeting the AI agent.
- Ingestion points: Data is pulled from AWS IAM credential reports, GitHub Organization/Repository API responses, and Okta user/application API endpoints as seen in
SKILL.md. - Boundary markers: The instructions and scripts do not implement specific boundary markers or "ignore instructions" delimiters to isolate external data from agent directives.
- Capability inventory: The skill possesses the capability to execute shell commands (bash), Python scripts, file system write operations, and network requests via CLI tools (
aws,gh,curl). - Sanitization: While data is structured using
jqandawk, there is no explicit security sanitization or validation performed on the resulting text before it is presented to or processed by the agent.
Audit Metadata