fix(acp): resolve reasoning_config for ACP and Feishu comment agents

`SessionManager._make_agent` and the Feishu doc-comment agent built their
`AIAgent` without `reasoning_config`, so `agent.reasoning_effort: none`
never reached those sessions: the transport applied its default effort,
which non-reasoning models such as gpt-4o-mini reject with HTTP 400 and
which silently re-enables thinking everywhere else. Both surfaces now go
through `hermes_constants.resolve_reasoning_config`, the same chokepoint
the CLI, gateway, TUI, cron and `hermes -p` already use, resolved against
the model the session actually runs so per-model overrides apply.

Ported from PR #85164 by @Chinmayrawat15 (oneshot hunk already on main).

Fixes #85153

acp_adapter/session.py

@@ -391,6 +391,7 @@ class SessionManager:
         from run_agent import AIAgent
         from hermes_cli.config import load_config
         from hermes_cli.runtime_provider import resolve_runtime_provider
+        from hermes_constants import resolve_reasoning_config
 
         config = load_config()
         model_cfg = config.get("model")
@@ -411,6 +412,10 @@ class SessionManager:
             "disabled_toolsets": list(disabled_toolsets) if disabled_toolsets is not None else None,
             "model": model or default_model,
             "cwd": cwd,
+            # Same chokepoint as the CLI/gateway/TUI/cron: without it ``agent.reasoning_effort: none`` never
+            # reaches an ACP session and the transport applies its default effort (a 400 on non-reasoning
+            # models). Resolved against the session's model so per-model overrides apply.
+            "reasoning_config": resolve_reasoning_config(config, model or default_model),
         }
         try:
             runtime = resolve_runtime_provider(