opencode-belweave

An opencode plugin that adds the Belweave Triad provider — an OpenAI-compatible aggregator gateway that routes to Claude, Gemini, GLM, GPT-OSS, and MiniMax models through a single endpoint.

Install the plugin and the provider + its model catalog appear automatically. No manual provider block in opencode.json is required.

Install

From npm

// opencode.json
{
  "$schema": "https://opencode.ai/config.json",
  "plugin": ["opencode-belweave"]
}

From a local path (development)

// opencode.json
{
  "$schema": "https://opencode.ai/config.json",
  "plugin": ["./packages/opencode-belweave"]
}

opencode runs on Bun, which loads the TypeScript entry directly — so no build step is needed while developing against a local path.

Configure your API key

Your key has the format trd_live_.... Provide it one of two ways:

Option A — environment variable:

export BELWEAVE_API_KEY=trd_live_xxxxxxxxxxxxxxxx

Option B — interactive login (stored in opencode's credential store):

Run opencode, then /connect → select Belweave Triad → paste your key.

Use a model

Run opencode, then /models → pick a Belweave Triad model, e.g. Claude Haiku 4.5 or GPT-OSS 20B.

Exposed models

16 models are exposed. Token limits are family-based estimates (used for opencode's context accounting); costs are set to 0 until the gateway's pricing is confirmed — override them in your own provider.belweave block if needed.

| Model id | Family | | --- | --- | | claude-haiku-4-5 | Claude | | claude-opus-4-1, claude-opus-4-5, claude-opus-4-6 | Claude | | claude-sonnet-4-5, claude-sonnet-4-6 | Claude | | gemini-3-1-flash-lite, gemini-3-1-pro, gemini-3-5-flash | Gemini | | glm-4-7-flash, glm-5 | GLM | | gpt-oss-120b, gpt-oss-20b | GPT-OSS | | gpt-oss-safeguard-120b, gpt-oss-safeguard-20b | GPT-OSS Safeguard (moderation) | | minimax-m2-5 | MiniMax |

Note: gpt-oss-safeguard-* are moderation / safety models and are exposed with minimal capabilities (text-only, no tool calling). Use them for classification / guardrails.

Overriding defaults

Anything you put under provider.belweave in opencode.json takes precedence over the plugin's defaults, including adding or tweaking individual models:

{
  "$schema": "https://opencode.ai/config.json",
  "plugin": ["opencode-belweave"],
  "provider": {
    "belweave": {
      "models": {
        "claude-opus-4-6": { "name": "Opus 4.6 (team)", "limit": { "context": 200000, "output": 16384 } }
      }
    }
  }
}

Requirements

• opencode >= 1.17.0

Releasing (maintainers)

Releases are published to npm automatically by GitHub Actions using npm trusted publishing (OIDC) — no npm token is stored anywhere. The publish workflow runs on every published GitHub Release; it typechecks, tests, builds, and then publishes with provenance.

To cut a release, run one of:

bun run release:patch   # 0.1.0 -> 0.1.1
bun run release:minor   # 0.1.0 -> 0.2.0
bun run release:major   # 0.1.0 -> 1.0.0

Each script bumps the version, pushes the tag, and creates a GitHub Release, which triggers the publish workflow.

One-time setup: trusted publishing requires the package to exist on npm first. After the initial bootstrap publish, configure the trusted publisher at npmjs.com → opencode-belweave → Settings → Trusted Publisher: provider GitHub Actions, repository aravhawk/opencode-belweave, workflow filename publish.yml.