openrouter-aider-orchestration

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt instructs embedding the OPENROUTER_API_KEY value directly into command arguments (e.g., "--openai-api-key", env["OPENROUTER_API_KEY"] and the example CLI with --openai-api-key "$OPENROUTER_API_KEY"), which requires the LLM/agent to include the secret verbatim in generated commands—an exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill invokes the OpenRouter API at https://openrouter.ai/api/v1 at runtime (via --openai-api-base) to obtain model outputs that directly drive agent prompts/behavior, and the orchestration relies on that external endpoint.