0 Fork 0

Files

T

Gemini Nexus Bot a905877c76 feat: bump version to 0.14.0

Includes all changes from this release cycle:
- Top header navigation (sidebar removed)
- Accounts screen decomposition and visual redesign
- Dev/prod environment isolation
- Usage Analytics with accurate averages, cost reporting, OpenRouter pricing
- Monitor moved to Usage with per-request token/cost tracking
- Clickable traffic logs with expandable JSON body viewer
- OpenCode IDE config in Quick Setup
- API Proxy reorganization (flattened tabs)
- CI/CD pipeline fixes (release config, Node version, pnpm alignment)
- Repository cleanup (~80MB of binaries removed)

2026-05-02 00:32:21 -03:00

3.4 KiB

Raw Permalink Blame History

Proxy Page — Feature Brainstorm

Generated from designer agent analysis of src/routes/proxy.tsx Date: 2026-04-22

Current State

The proxy page (src/routes/proxy.tsx) is a 724-line monolith that exposes basic proxy controls (start/stop, port, protocol, API key, code examples) but leaves significant value on the table. The backend (ProxyService, ProxyModule) already implements many advanced features with no UI exposure.

Proposed New Features

1. Live Traffic Monitor

Real-time SSE/WebSocket feed of proxied requests
Show request method, model, account used, latency, token count
Filter by status (success, rate-limited, error)
Inspired by: Request logging already exists in TrafficLogger

2. IDE Quick-Connect Panel

One-click copy of ready-to-use configurations for:
- Cursor (settings.json snippet)
- OpenCode (.env snippet)
- Windsurf (config.yaml snippet)
- Claude Code / CLI tools
Include the local IP, port, and API key pre-filled
QR code for mobile/quick access

3. API Playground

Built-in Swagger/OpenAPI UI or custom test panel
Pre-loaded with all supported endpoints:
- /v1/chat/completions (OpenAI-compatible)
- /v1/messages (Anthropic-compatible)
- /v1beta/models/...generateContent (Gemini-native)
Allow sending test requests and viewing formatted responses
Show exact curl equivalent for every request

4. Health Dashboard

Real-time metrics cards:
- Requests/minute
- Average latency (p50, p95, p99)
- Error rate %
- Active connections
- Circuit breaker status (open/closed/half-open)
Sparkline charts for the last 5 minutes
Color-coded status indicators

5. Model Availability Status

Live grid showing which models are currently available
Per-model indicators:
- Green: available
- Yellow: rate-limited (retry after)
- Red: unavailable / quota exhausted
Show last-used time and remaining quota (if known)
Collapse/expand by provider (Gemini, Anthropic)

6. Error Feed

Scrollable log of recent errors
Columns: timestamp, endpoint, model, account, error code, message
Click to expand full request/response
Filter by error type (rate limit, timeout, auth, network)
Export to JSON/CSV

7. Connection Visualizer

Diagram showing: IDE → Proxy → Account Pool → Upstream API
Animated flow dots representing live requests
Highlight currently active account
Show queue depth / waiting requests
Simple but effective for understanding routing

8. One-Click Diagnostics

"Run Diagnostics" button that performs:
- Check if proxy port is accessible
- Test each configured account
- Verify model mapping resolution
- Check upstream API reachability
- Validate API key format
Output: green checkmarks or detailed error report with fix suggestions

9. Session Stats

Summary of current session since proxy started:
- Total requests handled
- Total tokens proxied
- Unique accounts used
- Top 3 models used
- Uptime counter
Resettable per-session (not persisted)

Implementation Notes

Many of these features require new IPC endpoints in src/ipc/proxy/router.ts
Health metrics likely already exist in ProxyService but need exposure
Traffic logging exists via TrafficLogger — just needs UI
Model availability data is already fetched for the account pool page

3.4 KiB Raw Permalink Blame History