Problem
Hi — we represent A2Agent and would like to explore a technical and commercial integration with Codex Mate.
Codex Mate already manages provider configurations for Pi, OpenCode, KiloCode, Codex, and Claude Code. Its Pi configuration flow also supports custom OpenAI-compatible endpoints and remote model discovery.
However, A2Agent users currently need to configure the endpoint and protocol manually.
Proposed solution
We would like to verify A2Agent through Codex Mate’s existing provider architecture, initially focusing on the Pi integration:
- an optional A2Agent provider template using
https://api.a2agent.me/v1;
- automatic model discovery through
/v1/models;
- streaming and non-streaming chat;
- tool calls and structured error handling;
- preservation of existing API keys and custom configuration fields.
If useful, we can contribute the provider template, compatibility tests, and a tested setup guide. We can also discuss extending the configuration to OpenCode or KiloCode after the initial integration is verified.
Alternatives
If a built-in provider template is not preferred, we can contribute documentation and compatibility tests for the existing generic OpenAI-compatible configuration path.
Additional context
Separately, with your permission, we would like A2Agent to be listed in the README or project page under Supported Providers, Integrations, Partners, or another section you prefer:
A2Agent — an OpenAI- and Anthropic-compatible API gateway providing access to multiple LLM providers.
Would you be open to the technical integration and, separately, the link listing? Please let us know your preferred implementation scope, wording, link location, and whether a logo is acceptable.
For approved compatibility work, A2Agent can provide a private test key, technical support, and $20–$100 in non-cash API credits. These credits are not conditional on merging code, endorsement, referrals, or displaying the link.
If you’re interested, please feel free to contact us at saf096057@gmail.com.
Problem
Hi — we represent A2Agent and would like to explore a technical and commercial integration with Codex Mate.
Codex Mate already manages provider configurations for Pi, OpenCode, KiloCode, Codex, and Claude Code. Its Pi configuration flow also supports custom OpenAI-compatible endpoints and remote model discovery.
However, A2Agent users currently need to configure the endpoint and protocol manually.
Proposed solution
We would like to verify A2Agent through Codex Mate’s existing provider architecture, initially focusing on the Pi integration:
https://api.a2agent.me/v1;/v1/models;If useful, we can contribute the provider template, compatibility tests, and a tested setup guide. We can also discuss extending the configuration to OpenCode or KiloCode after the initial integration is verified.
Alternatives
If a built-in provider template is not preferred, we can contribute documentation and compatibility tests for the existing generic OpenAI-compatible configuration path.
Additional context
Separately, with your permission, we would like A2Agent to be listed in the README or project page under Supported Providers, Integrations, Partners, or another section you prefer:
Would you be open to the technical integration and, separately, the link listing? Please let us know your preferred implementation scope, wording, link location, and whether a logo is acceptable.
For approved compatibility work, A2Agent can provide a private test key, technical support, and $20–$100 in non-cash API credits. These credits are not conditional on merging code, endorsement, referrals, or displaying the link.
If you’re interested, please feel free to contact us at saf096057@gmail.com.