fix: increase curl timeout to 600s for long-running AI tasks
CURLOPT_TIMEOUT was 180s. Generating a consultation summary requires up to 10 AI iterations at ~20s each, regularly hitting the limit and causing a "שגיאה בתקשורת" error in the UI before Shira could respond. Raised to 600s. The shira-hermes agent runner now also enforces a 570s total budget so it returns a clean Hebrew message if the limit is hit, rather than the PHP connection being cut mid-response. Refs Task Master #2 Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
@@ -518,7 +518,7 @@ class SmartAssistantService
|
|||||||
CURLOPT_POSTFIELDS => $payload,
|
CURLOPT_POSTFIELDS => $payload,
|
||||||
CURLOPT_RETURNTRANSFER => true,
|
CURLOPT_RETURNTRANSFER => true,
|
||||||
CURLOPT_HTTPHEADER => $headers,
|
CURLOPT_HTTPHEADER => $headers,
|
||||||
CURLOPT_TIMEOUT => 180,
|
CURLOPT_TIMEOUT => 600,
|
||||||
CURLOPT_CONNECTTIMEOUT => 10,
|
CURLOPT_CONNECTTIMEOUT => 10,
|
||||||
]);
|
]);
|
||||||
|
|
||||||
|
|||||||
+2
-2
@@ -3,11 +3,11 @@
|
|||||||
"module": "SmartAssistant",
|
"module": "SmartAssistant",
|
||||||
"description": "Unified AI Assistant for Legal CRM — floating chat with case memory, office alerts, and AI Gateway integration",
|
"description": "Unified AI Assistant for Legal CRM — floating chat with case memory, office alerts, and AI Gateway integration",
|
||||||
"author": "klear",
|
"author": "klear",
|
||||||
"version": "2.8.0",
|
"version": "2.8.1",
|
||||||
"acceptableVersions": [
|
"acceptableVersions": [
|
||||||
">=8.0.0"
|
">=8.0.0"
|
||||||
],
|
],
|
||||||
"releaseDate": "2026-04-19",
|
"releaseDate": "2026-05-06",
|
||||||
"php": [
|
"php": [
|
||||||
">=8.1"
|
">=8.1"
|
||||||
]
|
]
|
||||||
|
|||||||
Reference in New Issue
Block a user