Maple AI is a TEE-based private inference provider using Confidential
Computing for end-to-end encryption with cryptographic attestations.
- Add maple-models.ts with model catalog and discovery
- Add MAPLE_API_KEY environment variable support
- Add interactive onboarding: `clawdbot onboard --auth-choice maple-api-key`
- Add configurable proxy URL (default: http://127.0.0.1:8080/v1)
- Add provider auto-registration when API key detected
- Add documentation at docs/providers/maple.md
Supported models: kimi-k2-thinking (default), gpt-oss-120b, deepseek-r1-0528,
qwen3-coder-480b, qwen3-vl-30b, llama-3.3-70b, gemma-3-27b
