[RaviTharuma]

Summary

Adds Pi-parity live model discovery to @omniroute/opencode-plugin so OpenCode picks up new OmniRoute models while the harness is running, and can force-refresh on demand.

Why

OpenCode previously only re-fetched the catalog on demand via modelCacheTtl after a provider/models hook. New models added in OmniRoute while a long-running OpenCode session was open could stay invisible until restart/TTL expiry + next hook. Pi already supports /omni sync + background discovery; OpenCode users need the same.

What changed

• autoSyncIntervalMs (default 300000 / 5 min, min 60000, 0 = off): background timer invalidates cache and re-fetches GET /v1/models (plus combos/enrichment when enabled) while OpenCode is running.
• omniroute_sync_models tool: force cache invalidate + refetch (OpenCode equivalent of Pi /omni sync).
• Command templates (OpenCode has no Pi-style slash-command API):
  • /omni-sync → asks the agent to call omniroute_sync_models
  • /omni-autosync → report current auto-sync / TTL status
• README docs for auto + force refresh paths.
• Plugin version 0.2.0 → 0.2.1.

Config example

{
  "plugin": [
    [
      "@omniroute/opencode-plugin",
      {
        "providerId": "omniroute",
        "baseURL": "https://or.example.com",
        "autoSyncIntervalMs": 300000
      }
    ]
  ]
}

Notes

• On-demand modelCacheTtl behavior is unchanged.
• Force path clears in-memory + optional disk snapshot caches, then repopulates from live OmniRoute endpoints.
• Based on release/v3.8.49 (includes management read-token plugin work from fix(opencode-plugin): support separate management read token #7885).

Test plan

• cd @omniroute/opencode-plugin && npm test (281 pass, 0 fail)
• npm run build (ESM + DTS success)
• tests/auto-sync.test.ts (10 pass) covering sanitize/default/clamp/0 disable
• Manual: run OpenCode with plugin, add a model in OmniRoute UI, wait ≤ autoSyncIntervalMs, confirm it appears without restart
• Manual: run /omni-sync (or ask agent to call omniroute_sync_models) and confirm immediate catalog refresh
• Manual: set autoSyncIntervalMs: 0 and confirm no background polling (TTL still works)

Closes #8106

[gemini-code-assist]

Caution

The consumer version of Gemini Code Assist on GitHub has been sunset. All code review activity has officially ceased.

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

[RaviTharuma]

Tracking issue: #8106

This PR implements Pi-parity live model discovery for OpenCode:

• background autoSyncIntervalMs (default 5m, min 60s, 0 = off)
• force tool omniroute_sync_models
• command templates /omni-sync + /omni-autosync

No credentials or host-specific secrets are included in the diff (examples only).

[diegosouzapw]

Merged into release/v3.8.49 — thanks @RaviTharuma! Validated via local merge-train (32-core) + a per-file discriminator confirming zero regressions vs the pure release tip. 🙏