plexr

Plexor CLI - LLM gateway, prompt optimization, and developer tools.

Installation

npm install -g @plexor-dev/plexr

Quick Start

# Set your API key
export PLEXOR_API_KEY="plx_user0_your_key_here"

# Send a prompt
plexr prompt "Explain recursion in one sentence"

# List available models
plexr models

# Check usage & savings
plexr usage

Commands

LLM Gateway

plexr prompt <text>

Send a prompt to Plexor's LLM gateway.

# Basic prompt
plexr prompt "What is the capital of France?"

# With specific model
plexr prompt "Write a haiku" --model gpt-4o

# With system prompt
plexr prompt "Translate to Spanish" --system "You are a translator"

# JSON output
plexr prompt "List 3 colors" --json

# With optimization (compression)
plexr prompt "Long prompt here..." --optimize

Options:

• -m, --model <model> - Model to use (default: claude-sonnet-4-20250514)
• -o, --optimize - Apply prompt compression
• --max-tokens <n> - Max response tokens (default: 4096)
• -t, --temperature <n> - Temperature 0.0-1.0 (default: 0.7)
• --system <prompt> - System prompt
• --json - Output raw JSON

plexr models

List all available models grouped by provider.

plexr models
plexr models --json

plexr usage

Show usage statistics and cost savings.

plexr usage
plexr usage --json

Token Sync (Claude MAX)

Sync your Claude MAX OAuth token to Plexor, enabling the platform to use your MAX subscription.

plexr sync

One-time sync of your Claude MAX token.

plexr sync
plexr sync --force  # Force sync even if unchanged

plexr status

Show token and sync status.

plexr status

plexr watch

Continuously watch for token changes and auto-sync.

plexr watch
plexr watch --interval 60  # Check every 60 seconds

plexr clear-cache

Clear the local sync cache.

plexr clear-cache

Configuration

Environment Variables

| Variable | Required | Description | Default | |----------|----------|-------------|---------| | PLEXOR_API_KEY | Yes | Your Plexor API key | - | | PLEXOR_API_URL | No | API endpoint | https://staging.api.plexor.dev |

Setting API Key

Linux/macOS:

echo 'export PLEXOR_API_KEY="plx_user0_your_key"' >> ~/.bashrc
source ~/.bashrc

Windows PowerShell:

[Environment]::SetEnvironmentVariable("PLEXOR_API_KEY", "plx_user0_your_key", "User")

Daemon Setup (Token Sync)

For automatic token sync, run plexr as a background service.

Linux (systemd)

# Install service
mkdir -p ~/.config/systemd/user
curl -o ~/.config/systemd/user/plexr.service \
  https://raw.githubusercontent.com/micoverde/plex-vc-fund-platform/main/scripts/plexr.service

# Enable and start
systemctl --user daemon-reload
systemctl --user enable --now plexr.service

# Check logs
journalctl --user -u plexr.service -f

macOS (launchd)

# Create plist at ~/Library/LaunchAgents/com.plexor.plexr.plist
# Then load:
launchctl load ~/Library/LaunchAgents/com.plexor.plexr.plist

Cron (simple)

# Sync every 30 minutes
*/30 * * * * PLEXOR_API_KEY=plx_xxx plexr sync >> /tmp/plexr.log 2>&1

Examples

Daily workflow

# Morning: check your usage
plexr usage

# Work: send prompts
plexr prompt "Review this code: $(cat myfile.py)"

# Quick questions
plexr prompt "What's the time complexity of quicksort?"

Scripting

# Pipe input
echo "Summarize this" | xargs plexr prompt

# JSON for parsing
plexr prompt "List 5 programming languages" --json | jq '.content'

# Batch processing
for file in *.txt; do
  plexr prompt "Summarize: $(cat $file)" >> summaries.txt
done

Supported Models

Run plexr models to see all available models. Includes:

• Anthropic: Claude Opus, Sonnet, Haiku (3.5, 4.0, 4.5)
• OpenAI: GPT-4o, GPT-4 Turbo, o1, o3
• DeepSeek: deepseek-chat, deepseek-reasoner
• Mistral: mistral-large, codestral
• Google: Gemini 2.0, 2.5, 3.0

License

MIT - Plexor Inc.