Set GPT-5.5 xhigh as agent default

[smilebank7]

Summary

• set every agent fallback model role to gpt-5.5(xhigh)
• update CLIProxyAPI env example and README setup snippet to match the default

Verification

• pnpm -F @openmath/agent typecheck
• pre-push hook: agent vitest 138 tests passed
• pre-push hook: math-engine pytest 20 tests passed
• CLIProxyAPI smoke request with model: gpt-5.5(xhigh) returned OK

---

Summary by cubic

Set gpt-5.5(xhigh) as the default model for all agent roles and updated CLIProxyAPI README/.env.example to match. This unifies defaults and enables stronger reasoning by default.

• Migration
  • If your router doesn’t support gpt-5.5(xhigh), pin LLM_MODEL/CLIPROXY_MODEL to an available model or upgrade your router.
  • For CLIProxyAPI, the examples now default to gpt-5.5(xhigh); no change needed unless you prefer a different model.

^{Written for commit 38af343. Summary will update on new commits.}

[cubic-dev-ai]

1 issue found across 3 files

Prompt for AI agents (unresolved issues)

Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.


<file name="packages/agent/src/config/models.ts">

<violation number="1" location="packages/agent/src/config/models.ts:5">
P2: Default model IDs now use a provider-specific alias format (`gpt-5.5(xhigh)`) that is passed unchanged to OpenAI, which expects the base model slug and separate reasoning-effort options.</violation>
</file>

<sub>Reply with feedback, questions, or to request a fix.

Re-trigger cubic</sub>

[cubic-dev-ai]

P2: Default model IDs now use a provider-specific alias format (gpt-5.5(xhigh)) that is passed unchanged to OpenAI, which expects the base model slug and separate reasoning-effort options.

Prompt for AI agents

Check if this issue is valid — if so, understand the root cause and fix it. At packages/agent/src/config/models.ts, line 5:

<comment>Default model IDs now use a provider-specific alias format (`gpt-5.5(xhigh)`) that is passed unchanged to OpenAI, which expects the base model slug and separate reasoning-effort options.</comment>

<file context>
@@ -1,12 +1,12 @@
-  solver: "gpt-4o",
-  objectiveLlm: "gpt-4o-mini",
+  intent: "gpt-5.5(xhigh)",
+  generator: "gpt-5.5(xhigh)",
+  constraintCritic: "gpt-5.5(xhigh)",
+  refiner: "gpt-5.5(xhigh)",
</file context>