Fully autonomous AI Agents system capable of performing complex penetration testing tasks
Go ai-agentsai-security-toolanthropicautonomous-agentsgolanggptgraphqlmulti-agent-systemoffensive-securityopen-sourceopenaipenetration-testingpenetration-testing-toolsreactsecurity-automationsecurity-testingsecurity-toolsself-hosted
fix(frontend): clear stale reasoning.maxTokens on model change (refs #288)
Sergey Kozyrenkoe18169f1a63c
diff --git a/frontend/src/pages/settings/settings-provider.tsx b/frontend/src/pages/settings/settings-provider.tsxindex 4c6b3ceb..d9568171 100644--- a/frontend/src/pages/settings/settings-provider.tsx+++ b/frontend/src/pages/settings/settings-provider.tsx@@ -1544,7 +1544,7 @@ function SettingsProvider() {);// Reset reasoning on model change: adaptive-only models lock- // to adaptive, others clear the now-stale mode/effort.+ // to adaptive, others clear the now-stale mode/effort/budget.setValue(`agents.${agentKey}.reasoning.mode` as const,option?.reasoning?.mode === ModelReasoningMode.AdaptiveOnly@@ -1552,6 +1552,10 @@ function SettingsProvider() {: null,);setValue(`agents.${agentKey}.reasoning.effort` as const, null);+ setValue(+ `agents.${agentKey}.reasoning.maxTokens` as const,+ null,+ );}}options={availableModels}placeholder="Select or enter model name"