Wizard Proxy

{{ logsAutoRefreshInterval ? `Auto (${logsAutoRefreshInterval >= 60000 ? (logsAutoRefreshInterval/60000) + 'min' : (logsAutoRefreshInterval/1000) + 's'})` : 'Refresh' }}

Refresh

Sync OpenRouter

Refresh

Channels {{ (modelsChannelFilter || channelRailAttentionOnly) ? filteredChannelRailServices.length + ' / ' : '' }}{{ settingsServices.length }} {{ channelOverview.enabled }} enabled accounts

Sort

No matching channels.

{{ isChannelDisabled(currentSettingsService.id) ? 'Off' : 'On' }}

Share live usage

{{ getServiceActiveConfigCount(currentSettingsService.id) }} / {{ getServiceConfigCount(currentSettingsService.id) }} enabled {{ formatNumber(currentSettingsRecentSummary.totalRequests || 0) }} requests 5m {{ getServiceOverviewRate(currentSettingsService.id) }} success {{ formatNumber(getServiceRoutingTotal(currentSettingsService.id)) }} routed

Disabled {{ getAuthTypeLabel(config.authType) }} w{{ getConfigStats(currentSettingsService.id, config.name)?.weight ?? config.weight }} · {{ calculateTrafficPercent(currentSettingsService.id, config.name) }}% Reserved · {{ config.allowedApiKeyIds.length }} {{ config.allowedApiKeyIds.length === 1 ? 'key' : 'keys' }} {{ getConfigIdentityHint(config, currentSettingsService.id) }}

Plan usage

No plan usage for this auth mode

Live 5m

{{ getWindowSummary(currentSettingsService.id, config.name, '5m').total_requests }} req {{ formatSummarySuccessRate(getWindowSummary(currentSettingsService.id, config.name, '5m')) }} ok {{ getWindowSummary(currentSettingsService.id, config.name, '5m').failure }} err {{ configLatencyText(currentSettingsService.id, config.name) }}

No requests in the last 5m

{{ group.model }} {{ describeFailureGroup(group).slice(group.model.length + 3) }} ×{{ group.count }}
+{{ configFailureGroups(currentSettingsService.id, config.name).length - 2 }} more

Models

Any (general)

By model Calendar week / month (UTC)

Model	This week	This month	At this pace
{{ model.model }}	{{ model.week_requests.toLocaleString() }} req · {{ formatTokensK(model.week_tokens) }} tok	{{ model.month_requests.toLocaleString() }} req · {{ formatTokensK(model.month_tokens) }} tok	{{ modelRunwayText(model) }}

Protocol · Health · Diagnostics

Protocol / Upstream

Wire protocol and target host

Protocol {{ getConfigProtocolLabel(config, currentSettingsService.id) }}

Upstream {{ getConfigUpstream(config, currentSettingsService.id) }}

Health

Recent success ratio and activity

{{ getHealthPercent(getConfigStats(currentSettingsService.id, config.name)) }} OK {{ getConfigStats(currentSettingsService.id, config.name)?.total_success || 0 }}/{{ getConfigStats(currentSettingsService.id, config.name)?.total_requests || 0 }}

{{ getWindowSummary(currentSettingsService.id, config.name, '24h').total_requests }} / 24h {{ getConfigStats(currentSettingsService.id, config.name)?.last_updated ? formatLastUpdated(getConfigStats(currentSettingsService.id, config.name)?.last_updated) : 'No activity' }}

Req 24h {{ getWindowSummary(currentSettingsService.id, config.name, '24h').total_requests }}

OK Rate {{ formatSummarySuccessRate(getWindowSummary(currentSettingsService.id, config.name, '24h')) }}

Routed {{ getRoutingStats(currentSettingsService.id, config.name)?.count || 0 }}

Last {{ getLastResultLabel(currentSettingsService.id, config.name) }}

Request Volume

Last 24 hours by configuration

Latency

Recent and hourly distribution

Retired {{ retiredSettingsConfigs.length }} Kept on the books, not served, not sampled. Re-list one to bring it back (disabled).

{{ config.name }} {{ config.id }} {{ getAuthTypeLabel(config.authType) }} {{ getConfigIdentityHint(config, currentSettingsService.id) }} {{ getConfigModelList(config).join(', ') }}

Re-list

{{ activePromptCount }} / {{ PROMPT_SERVICES.length }} services with prompts Applies to new requests

{{ activePromptService.label }} prompt

{{ activePromptService.description }} · {{ activePromptService.field }}

{{ !customPromptsLoaded ? 'Not loaded' : isCustomPromptDirty(selectedPromptService) ? 'Unsaved changes' : 'Saved' }}

Instructions

Plain text · empty means no injected prompt{{ customPrompts[selectedPromptService].prompt.length.toLocaleString() }} characters

Clear draft

How prompt injection works

Prompts apply to all requests for the selected service. Claude uses the top-level system field; Codex uses instructions; Gemini and Grok use a system message. Draft changes do not affect traffic until published. Clearing the editor also requires publishing.

New Key

Key	Status	Requests	Traffic	Models	Last Used
{{ keyEntry.name }} ID {{ (keyEntry.id \|\| '').slice(0, 8) \|\| '—' }}	{{ keyEntry.active ? 'Active' : 'Disabled' }}	{{ formatNumber(keyEntry.requestCount \|\| 0) }}	In {{ formatTokensK(keyEntry.inputTokens \|\| 0) }} Out {{ formatTokensK(keyEntry.outputTokens \|\| 0) }} Total {{ formatTokensK((keyEntry.inputTokens \|\| 0) + (keyEntry.outputTokens \|\| 0)) }}	{{ formatAllowedModels(keyEntry) }}	{{ keyEntry.lastUsedAt ? formatRelativeTime(keyEntry.lastUsedAt) : 'Never' }}	Copy Edit

No Access Keys

Create your first access key to enable client authentication.

Create Key

Name

Status

Rate Limits

RPM

TPM

Quota

Limit

Type

Reset Period

Expires At

Allowed Models

All Models

Select Models

No models in Pricing yet

If not configured, this key can use every model.

Allowed Sites

All Sites

Select Sites

A friendly name to identify this key

All Models (default)

Select specific models

No models in Pricing yet

Leave as All Models to avoid restrictions.

All Sites (default)

Select specific sites

No sites configured yet

Key Created Successfully

Make sure to copy this key now. You won't be able to see it again!

Your New Access Key

{{ newlyCreatedKey.key }}

Copy

Request History

{{ historyPagination.total }} logs

Prompts

Success {{ historySummary.success }}

Error {{ historySummary.failure }}

Service Session Model Route

{{ log.status_code || '—' }} {{ getLogServiceType(log) }} — {{ routedFromLabel(log) }}{{ log.model || '—' }} {{ log.thinking_level }} {{ log.channel || '—' }}{{ formatRouteTag(log.route_tag) }}

{{ log.duration_ms ? (log.duration_ms >= 1000 ? (log.duration_ms / 1000).toFixed(1) + 's' : log.duration_ms + 'ms') : '—' }} {{ timeAgo(log.timestamp) }}

⚠ {{ log.error_type }} {{ log.method }} {{ truncatePath(log.path) }} {{ log.agent || '—' }} {{ logTokenSummary(log) }} {{ logCostLabel(log) }} {{ log.api_key_name }} {{ log.client_ip }}

Conversation · session_id

`{{ selectedSessionId || '—' }}`

cwd {{ sessionCwd }}

session {{ sessionTrace.session_id || selectedSessionId }} · {{ sessionTrace.span_count || sessionTrace.total || 0 }} spans · {{ formatTokensK(sessionTrace.total_tokens || 0) }} tokens · {{ ((sessionTrace.total_duration_ms || 0) / 1000).toFixed(1) }}s · {{ (sessionTrace.models || []).join(', ') }}

Refresh

Open full detail

No session selected

Spans {{ (sessionTrace.spans || []).length }}{{ (sessionTrace.span_count || sessionTrace.total) ? (' / ' + (sessionTrace.span_count || sessionTrace.total)) : '' }}

Select a span

Overview metrics

Requests, tokens, estimated cost, and pricing coverage

Requests {{ formatNumber(insightsOverview?.totals?.requestCount || 0) }} successful

Tokens {{ formatTokensK(insightsOverview?.totals?.totalTokens || 0) }} {{ formatTokensK(insightsOverview?.totals?.inputTokens || 0) }} in · {{ formatTokensK(insightsOverview?.totals?.outputTokens || 0) }} out

Est. API Cost {{ formatCostUsd(insightsOverview?.totals?.costUsd || 0) }} {{ formatCostUsd(insightsOverview?.totals?.avgCostUsd || 0) }} / req · OpenRouter rates

$ / 1M Tokens {{ formatCostUsd(insightsOverview?.totals?.costPerMTokens || 0) }} blended · priced tokens only

Pricing Coverage {{ (insightsOverview?.totals?.pricingCoverage || 0).toFixed(1) }}% {{ insightsOverallModels.length }} models shown · unpriced excluded from rate

{{ insightsError }} · Showing last successful result.

Loading usage…

No usage data for this window.

Channel usage

Provider and model costs · OpenRouter rates

{{ formatTokensK(insightsOverview?.totals?.totalTokens || 0) }} · {{ formatCostUsd(insightsOverview?.totals?.costUsd || 0) }}

ProviderChannelsRequestsTokensCostCost share

ChannelWindow creditsPlan usage

Model ranking

Sorted by tokens · estimated from OpenRouter rates

# Model Requests Tokens Cost Cost share

Agent ranking

Sorted by tokens · estimated from OpenRouter rates

# Agent Requests Tokens Cost Cost share

Model metrics

Usage, estimated cost, and users for this model

{{ insightsError }} · Showing last successful result.

Loading model usage…

Agent share

Who calls this model

Used by agents

Client agents calling this model

{{ insightsModelAgents.length }} agents

{{ item.value }} {{ formatTokensK(item.totalTokens) }}

{{ item.requestCount }} req {{ item.requestShare.toFixed(1) }}% {{ formatCostUsd(item.costUsd) }} last {{ item.lastSeen.replace('T', ' ').substring(0, 16) }}

No agent data

Used by keys

API keys calling this model

{{ insightsModelKeys.length }} keys

{{ item.value }} {{ formatTokensK(item.totalTokens) }}

{{ item.requestCount }} req {{ item.requestShare.toFixed(1) }}% {{ formatCostUsd(item.costUsd) }} last {{ item.lastSeen.replace('T', ' ').substring(0, 16) }}

No key data

Routing configs

Upstream config footprint for this model

{{ insightsModelConfigs.length }} configs

{{ item.value }} {{ item.requestCount }} req

{{ item.requestShare.toFixed(1) }}% {{ formatTokensK(item.totalTokens) }} {{ formatCostUsd(item.costUsd) }}

No usage data for this model and window.

Agent metrics

Usage, estimated cost, and coverage for this agent

{{ insightsError }} · Showing last successful result.

Loading agent usage…

Model share

Token distribution for this agent

Models used

Which models this agent requests

{{ insightsAgentModels.length }} models

{{ item.value }} {{ item.costUsd ? formatCostUsd(item.costUsd) : formatTokensK(item.totalTokens) }}

{{ item.requestCount }} req {{ item.requestShare.toFixed(1) }}% {{ formatTokensK(item.totalTokens) }}

Keys used

Which API keys this agent uses

{{ insightsAgentKeys.length }} keys

{{ item.value }} {{ formatTokensK(item.totalTokens) }}

{{ item.requestCount }} req {{ item.requestShare.toFixed(1) }}% {{ formatCostUsd(item.costUsd) }} last {{ item.lastSeen.replace('T', ' ').substring(0, 16) }}

Key activity

Hourly request heatmap per key

{{ insightsAgentKeyActivity[0].buckets[0].bucket.replace('T', ' ') }} {{ insightsAgentKeyActivity[0].buckets[insightsAgentKeyActivity[0].buckets.length - 1].bucket.replace('T', ' ') }}

No usage data for this agent and window.

{{ selectedLog.service }} · {{ formatChannelName(selectedLog.channel) }} {{ selectedLog.route_tag }} {{ routedFromLabel(selectedLog) }} → {{ selectedLog.model }} · {{ selectedLog.model }} · effort {{ selectedLog.thinking_level }} · {{ selectedLog.duration_ms }}ms · {{ selectedLog.timestamp }}

Session {{ String(selectedLog.session_id).slice(0, 8) }}

Routing trigger: {{ selectedLog.route_reason }}

💬 User Message

Client → Proxy

Host: 127.0.0.1:3211

Path: {{ selectedLog.path }}

Proxy → Upstream

N/A

Host: {{ getUpstreamHost(selectedLog.target_url) }}

Path: {{ getUpstreamPath(selectedLog.target_url) }}

Not forwarded

Request

{{ selectedLog.path }}

Copy

Client: {{ selectedLog.client_ip }} ID: {{ selectedLog.request_id }}

Headers ({{ Object.keys(selectedLog.original_headers || {}).length }})

Copy

No headers

{{ key }} {{ formatHeaderDisplayValue(key, selectedLog.original_headers[key]) }}

Body ({{ formatBytes(selectedLog.original_body_length) }})

Truncated

Copy

Format

No body

Response {{ selectedLog.status_code }}

Status: {{ selectedLog.status_code }} {{ getStatusText(selectedLog.status_code) }} Size: {{ formatBytes(selectedLog.response_bytes) }} Duration: {{ selectedLog.duration_ms }}ms

Headers ({{ Object.keys(selectedLog.response_headers || {}).length }})

Copy

No headers

{{ key }} {{ formatHeaderDisplayValue(key, selectedLog.response_headers[key]) }}

Body ({{ formatBytes(selectedLog.response_content_length) }})

Truncated

Copy

Format

No response

Request was not forwarded to upstream

Realtime requests

Claude is connected

Claude is not connected

Codex is connected

Codex not connected

connect

disconnect

No real-time request yet

Serve path state time consuming operate

{{ request.service }}{{ request.channel ? ` [${request.channel}]` : '' }} {{ request.path }}

{{ request.displayDuration }}ms

Details

Request history

Total {{ (historyPagination && historyPagination.total) || 0 }} logs (current page {{ filteredLogs.length }} logs, total {{ (historyPagination && historyPagination.overallTotal) || 0 }} logs, keep the most recent {{ systemConfig.logLimit }} logs)

Success {{ historySummary.success }}

Error {{ historySummary.failure }}

refresh

time Serve URL state time consuming Usage operate

{{ allLogsLoading ? 'Loading log data...' : ((allLogs.items && allLogs.items.length === 0) ? 'No log data' : 'No matching log') }}

{{ formatTimestamp(log.timestamp) }} {{ formatServiceWithChannel(log.service, log.channel) }} {{ formatMethodWithURL(log.method, log.path) }}

{{ log.duration_ms ? `${log.duration_ms}ms` : '-' }} {{ formatUsageSummary(log.usage, log.service) }}

Details

{{ selectedRealtimeRequest.displayDuration }}ms

real-time response streaming

{{ selectedRealtimeRequest.responseText }}

No response data yet...

Request header

{{ stringifyHeadersForDisplay(selectedRealtimeRequest.request_headers) }}

{{ healthCounts[id] || 0 }} {{ meta.label.toLowerCase() }}

{{ healthState.error ? 'Upstream health is unavailable right now.' : 'No requests in this window.' }}

Proxy transitions in window

{{ healthEvents.length }} lines from server_log · also drawn as markers

{{ upstreamHealth.formatTime(e.time) }} {{ e.kind_label }} {{ e.message }}

Routes {{ (routingFilter || routingAttentionOnly) ? routingVisibleCount + ' / ' : '' }}{{ routingSummary.routes }} {{ routingSummary.enabled }} on checking channels…

+ Add route

No route matches "{{ routingFilter }}".

{{ routingSaveNote }} ⌘S

{{ pricingSyncSummary.matched }}/{{ pricingSyncSummary.total }} matched Quota and Kiro multipliers are manual overrides

Pricing and quota calculation

Models are discovered from configured services; USD prices sync from OpenRouter. API-key quotas use raw total tokens × quota multiplier. Raw token counts are recorded separately.

refreshed {{ serverLogsRelative(serverLogsFetchedAt) }}

Server log mirror is not available

{{ serverLogsUnavailable }} · the per-slot file ~/.evot/run/unified_proxy-<port>.log remains the authority.

Lines	Level	Condition	Channels	Requests	First	Last
Nothing in the last {{ serverLogsWindow }}
{{ sh.lines }}	{{ sh.errors ? (sh.warnings ? 'mixed' : 'error') : 'warning' }}	{{ sh.shape }}	{{ sh.services.join(', ') \|\| '—' }}	{{ sh.requests \|\| '—' }}	{{ serverLogsClock(sh.since) }}	{{ serverLogsRelative(sh.last) }}

Time	Level	Channel	Message	Request
No lines match
{{ serverLogsClock(row.ts) }}	{{ row.level.toLowerCase() }}	{{ row.service }} / {{ row.config_name }}—	{{ serverLogsHead(row.message) }}	{{ row.request_id.slice(0, 8) }}—
{{ row.message }} logger {{ row.logger }} instance {{ row.instance }} · {{ row.host }}

Request

copy id

open in Logs

Hops request history

No request-history row yet — the request failed before it was recorded, or the row is still on its way.

{{ h.hop }} {{ h.status_code || '—' }} {{ h.service }} / {{ h.config_name }} {{ h.model }}

Log lines {{ serverLogsTrace.lines.length }}, oldest first

{{ serverLogsClock(l.ts) }} {{ l.level.toLowerCase() }} {{ serverLogsHead(l.message) }} — {{ l.logger }}

Loading…

{{ currentSettingsService.label }}

Protocol / Upstream

Health

Request Volume

Latency

{{ activePromptService.label }} prompt

No Access Keys

Key Created Successfully

Request History

`{{ selectedSessionId || '—' }}`

Overview metrics

Channel usage

Model ranking

Agent ranking

Model metrics

Agent share

Used by agents

Used by keys

Routing configs

Agent metrics

Model share

Models used

Keys used

Key activity

Proxy API Key (Claude / Codex / Gemini)

Realtime requests

Request history

real-time response streaming

Request header

Proxy transitions in window

Failover order

Server log mirror is not available

Conditions

Lines

{{ serverLogsTrace.request_id }}

Hops request history

Log lines {{ serverLogsTrace.lines.length }}, oldest first

Protocol / Upstream

Health

Request Volume

Latency

No Access Keys

Key Created Successfully

Request History

{{ selectedSessionId || '—' }} Copy

Overview metrics

Channel usage

Model ranking

Agent ranking

Model metrics

Agent share

Used by agents

Used by keys

Routing configs

Agent metrics

Model share

Models used

Keys used

Key activity

Proxy API Key (Claude / Codex / Gemini)

Realtime requests

Request history

real-time response streaming

Request header

Proxy transitions in window

Failover order

Server log mirror is not available

Conditions

Lines

Hops request history

Log lines {{ serverLogsTrace.lines.length }}, oldest first

`{{ selectedSessionId || '—' }}`