update

backend/agents.py

@@ -13,9 +13,11 @@ import signal
 import tempfile
 import time
 import urllib.request
+from contextlib import asynccontextmanager
 from pathlib import Path
 
-from config import PROVIDERS, DEFAULT_PROVIDER, MAX_CONCURRENT_AGENTS, MAX_CONCURRENT_INTERACTIVE
+from config import (PROVIDERS, DEFAULT_PROVIDER, MAX_CONCURRENT_AGENTS,
+                    MAX_CONCURRENT_AGENTS_PER_TOPIC, MAX_CONCURRENT_INTERACTIVE)
 
 log = logging.getLogger("creator.agents")
 
@@ -43,6 +45,24 @@ def _scope_cancelled(agent_key: str) -> bool:
 # does not count against the agent timeout.
 _batch_sem = asyncio.Semaphore(MAX_CONCURRENT_AGENTS)
 _interactive_sem = asyncio.Semaphore(MAX_CONCURRENT_INTERACTIVE)
+# Per-topic caps (lazily created): each topic gets its own batch semaphore of size
+# MAX_CONCURRENT_AGENTS_PER_TOPIC, nested INSIDE the global _batch_sem.
+_topic_sems: dict[str, asyncio.Semaphore] = {}
+
+
+@asynccontextmanager
+async def _batch_gate(scope: str | None):
+    """Acquire a batch slot: per-topic semaphore FIRST, then the global one. The order matters —
+    a waiter holds only its (per-topic) slot while queueing for the global cap, so a saturated topic
+    never blocks other topics on the global semaphore. scope=None → global cap only."""
+    topic_sem = _topic_sems.setdefault(scope, asyncio.Semaphore(MAX_CONCURRENT_AGENTS_PER_TOPIC)) if scope else None
+    if topic_sem is None:
+        async with _batch_sem:
+            yield
+    else:
+        async with topic_sem:
+            async with _batch_sem:
+                yield
 
 # Serialize OpenCode starts: processes starting simultaneously collide on the
 # internal session DB ("database is locked", exit after <1s). The short
@@ -116,6 +136,7 @@ async def run_agent(
     role: str = "fast",
     capabilities: str = "none",
     lane: str = "batch",
+    scope: str | None = None,
 ) -> tuple[int, str, str]:
     if _scope_cancelled(agent_key):  # before queueing: don't even enter the queue
         return 1, "", "cancelled"
@@ -123,8 +144,8 @@ async def run_agent(
         return 1, "", f"Unknown provider: {provider}"
     if shutil.which(PROVIDERS[provider]["cli"]) is None:
         return 1, "", f"CLI '{PROVIDERS[provider]['cli']}' not installed (provider: {provider})"
-    sem = _interactive_sem if lane == "interactive" else _batch_sem
-    async with sem:
+    gate = _interactive_sem if lane == "interactive" else _batch_gate(scope)
+    async with gate:
         if _scope_cancelled(agent_key):  # after the acquire: cancelled in the queue → no spawn
             return 1, "", "cancelled"
         if PROVIDERS[provider]["cli"] == "opencode":