openrouter-common-errors
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill interacts with official OpenRouter API domains (
openrouter.ai) for legitimate diagnostic tasks, including verifying API keys, checking credit balances, and testing model availability. These operations are consistent with the skill's stated purpose. - [INDIRECT_PROMPT_INJECTION]: The skill provides templates for processing user-supplied message data that is eventually sent to an external LLM provider, creating a standard attack surface for indirect prompt injection.
- Ingestion points: User-controlled content enters the system via the
messageslist in the diagnostic and error-handling functions located inSKILL.mdandreferences/examples.md. - Boundary markers: The implementation does not include specific boundary delimiters or safety instructions to prevent the model from following commands embedded within the user messages.
- Capability inventory: The skill is configured to use network-capable tools including
curl,python, andnodeto communicate with the provider API. - Sanitization: The skill includes a
validate_before_sendfunction that performs schema validation (ensuring required fields like 'role' and 'content' are present) and estimates token usage to avoid context length errors.
Audit Metadata