examples: make the agent-server trial timeout configurable

examples/swe-agent/README.md

@@ -48,6 +48,19 @@ generous — agentic trials routinely run past an hour, and a short timeout kill
 them mid-episode. Verify `http://<agent-server>:30000/health` before launching
 Miles.
 
+The two per-trial timeouts must be ordered. `--agent-timeout` is the authoritative
+one: when it fires, the agent server ends the trial and frees its sandbox. The
+rollout client applies a second ceiling, `AGENT_TRIAL_TIMEOUT` (default 7200
+seconds), which has to stay above `--agent-timeout`. If the client gives up first,
+the trial is recorded as aborted while the agent server keeps running it, so the
+sandbox and its `--max-concurrent` slot stay busy for the remaining difference, and
+the aborted sample takes its whole GRPO group down with it. Raise it through the
+launcher's generic env-var hook:
+
+```bash
+python examples/swe-agent/run.py ... --extra-env-vars 'AGENT_TRIAL_TIMEOUT=10800'
+```
+
 If the trainer reaches the agent server through a proxy or an in-cluster service
 rather than directly, point `--agent-server-url` at that stable name rather than
 an ephemeral pod address. The rollout client enables TCP keepalive probes so

examples/swe-agent/swe_agent_function.py

@@ -23,9 +23,17 @@
 
 logger = logging.getLogger(__name__)
 
+# Backstop for an unreachable agent server; its own --agent-timeout should fire first.
+_DEFAULT_AGENT_TRIAL_TIMEOUT_S = 7200
+
 _agent_server_client: httpx.AsyncClient | None = None
 
 
+def _agent_trial_timeout_s() -> int:
+    """Per-trial ceiling for the agent-server call, overridable via AGENT_TRIAL_TIMEOUT."""
+    return int(os.environ.get("AGENT_TRIAL_TIMEOUT", _DEFAULT_AGENT_TRIAL_TIMEOUT_S))
+
+
 def _get_agent_server_client() -> httpx.AsyncClient:
     """Return a client whose long-running requests survive idle network paths."""
     global _agent_server_client
@@ -102,13 +110,14 @@ async def run(
     if session_server_instance_id is not None:
         request["session_server_instance_id"] = session_server_instance_id
 
+    trial_timeout_s = _agent_trial_timeout_s()
     try:
         response = await asyncio.wait_for(
             _post_agent_server(f"{agent_server_url}/run", request),
-            timeout=3600,  # 1 hour max per trial
+            timeout=trial_timeout_s,
         )
     except asyncio.TimeoutError:
-        logger.error("Agent server call timed out after 3600s")
+        logger.error(f"Agent server call timed out after {trial_timeout_s}s")
         return None
     except asyncio.CancelledError:
         logger.warning("Agent server call cancelled (sibling task failure?)")