Add OpenCode agent backend

[simbo1905]

Summary

Adds an optional opencode agent backend, with its provider/model selection passed as ordinary role arguments.

Also included:

• an opt-in, smoke-only role instruction profile for bounded diagnostics;
• SWARMFORGE_CONFIG selection of an alternate project-local configuration; and
• a handoff-daemon detachment fix so the daemon follows its explicit stop-file lifecycle rather than being torn down with the terminal bridge.

Closes #39.

Scope change

This PR previously also carried the Mistral Vibe backend. Vibe has been split out into #46 and its own PR so that each backend can be reviewed independently. This branch no longer references Vibe at all.

Validation

bb test on this branch, compared against unmodified main on the same host:

Branch	Tests	Assertions	Failures	Errors
main	24	87	0	2
this branch	26	94	0	2

The two errors are close-swarm-kills-tmux-sessions-and-stops-daemon and swarmforge-detects-nonzero-pane-base-index, which both require tmux and error identically on unmodified main in this tmux-less host environment. No new failures are introduced. The full suite previously passed in a mountless aarch64 Lima guest.

New coverage:

• opencode-launch-command-passes-model-and-initial-prompt
• swarmforge-role-only-instruction-profile-is-explicit-and-bounded

Scope and security

No credentials, telemetry, or benchmark artifacts are included. OpenCode must be installed and authenticated by the runtime owner.

[simbo1905]

Hey @unclebob

This one is a little half baked so far as is making some arbitrary decisions such as it picking run this open model to keep my token costs down while I am trying things out myself

"--model opencode-go/qwen3.7-plus --auto"

I think there is an case to be made that I split out mistral.ai as its a mono provider backend; so it has its own standard model defaults. So that can probably get onto main faster.

On the question of what is the default provider for OpenCode (Go at $10 a month, or Zen with pay-as-you-go), and the default models. There are Deepseek, Kimi and others hosted in US, EU and Singapore on OpenCode. That makes things more complex but a ton of fun https://opencode.ai/docs/zen/#pricing

It could be fun to come up with some swarm-forge benchmark tasks and to then different mix and matches of open and proprietary models perhaps a "Swarm Bench"...

Give me a thumbs up if you want me to pursue this.

Simon