feat: Add OpenAI agents config

sources/agents/openai_completion.toml

@@ -0,0 +1,30 @@
+schema_version = 1
+
+extends     = "OpenAI Base Chat"
+name        = "OpenAI Completion"
+description = "OpenAI GPT-5.4 mini — code completion using the <code_context> chat format over Chat Completions. reasoning_effort=none for low latency. GPT-5 models reject 'temperature' AND 'stop' on Chat Completions, so completion cannot be pinned to t=0 and has no stop sequence — it relies on the prompt + max_completion_tokens to stay short."
+
+model = "gpt-5.4-mini"
+tags  = ["completion", "openai", "cloud"]
+
+system_prompt = """
+{%- if language == "qml" %}{{ read_file(":/roles/code-completion-qml.md") }}
+{%- else if language == "c-like" %}{{ read_file(":/roles/code-completion-c-like.md") }}
+{%- else %}{{ read_file(":/roles/code-completion.md") }}
+{%- endif %}
+{{ read_file(":/tasks/code-completion.md") }}"""
+
+[body]
+max_completion_tokens = 1024
+reasoning_effort      = "none"
+messages = """
+[
+  {% if existsIn(ctx, "system_prompt") %}
+  { "role": "system", "content": {{ tojson(ctx.system_prompt) }} },
+  {% endif %}
+  {
+    "role": "user",
+    "content": {{ tojson("Here is the code context with insertion points:\\n<code_context>\\n" + ctx.prefix + "<cursor>" + ctx.suffix + "\\n</code_context>") }}
+  }
+]
+"""