llm-router

[Skill Scanner] Natural language instruction to download and install from URL detected This skill/documentation is consistent with a legitimate commercial API gateway that centralizes access to many LLM providers using one API key. There is no direct evidence of obfuscated or malicious code in the provided text. The primary security concern is architectural: the gateway (api.aisa.one) receives all user prompts, images, and the AISA_API_KEY, meaning sensitive data and credentials are concentrated at a third-party endpoint rather than going directly to vendor APIs. That design is not inherently malicious but increases risk and requires explicit trust and clear privacy/security guarantees from the gateway operator. Recommendation: treat this as suspicious from a supply-chain/trust perspective unless you independently verify the operator, review their privacy/retention/logging and security practices, and consider whether direct vendor integrations (or client-side encryption/proxying) are preferable for your threat model. LLM verification: This skill is a documentation-only description of a third-party LLM gateway (api.aisa.one) that requires a single API key and routes all prompts, images, and function calls through that service. There is no explicit malicious code in the supplied file, but the design centralizes credentials and sensitive data at a third-party endpoint. If a developer expects direct calls to official provider APIs, this gateway model is a mismatch and introduces a supply-chain/trust risk (credential capture, logg