workos-api-admin-portal
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION] (LOW): Indirect Prompt Injection vulnerability surface detected.
- Ingestion points:
SKILL.mdcontains a 'Step 1' instruction for the agent to 'WebFetch' four specific documentation URLs fromworkos.comto get 'latest implementation details'. - Boundary markers: Absent. There are no instructions or delimiters provided to ensure the agent ignores malicious instructions that could theoretically be embedded in the retrieved web content.
- Capability inventory: The skill enables the agent to construct and execute
curlcommands and Node.js scripts (via@workos-inc/node) to interact with the WorkOS API. - Sanitization: Absent. Content fetched from the URLs is expected to be processed directly to inform agent behavior.
- [EXTERNAL_DOWNLOADS] (SAFE): The skill references the official WorkOS Node.js SDK (
@workos-inc/node). While this is an external dependency, it is the standard library for the service described and the reference occurs within documentation examples rather than an automated installation script. - [CREDENTIALS_UNSAFE] (SAFE): Example code includes a hardcoded API key (
sk_test_1234567890abcdef). This is identified as a dummy/placeholder test key used for educational purposes and does not represent a leak of live production credentials.
Audit Metadata