RunAPI CLI Skill - Setup Guide¶

Unified AI media API CLI for agents by runapi-ai. Consolidates AI image generation, video creation, music/audio synthesis, and model API access under a single CLI - eliminating the need for separate integration skills per provider.

What It Provides¶

AI Image Generation - DALL-E, Stable Diffusion, Flux, Midjourney (via API)
AI Video - Runway, Pika, Kling, HeyGen
AI Music/Audio - Suno, Udio, ElevenLabs TTS
AI Model APIs - OpenAI, Anthropic, Google, DeepSeek, local Ollama
Job Queue - async job submission with status polling
Cost Tracking - per-provider usage and cost breakdown

Installation¶

# Install via skills.sh
npx skills add https://github.com/runapi-ai/cli-skill

# Manual clone
git clone https://github.com/runapi-ai/cli-skill.git ~/.hermes/skills/runapi-cli

Configuration¶

runapi:
  providers:
    openai:
      api_key: "${OPENAI_API_KEY}"
      models: ["dall-e-3", "gpt-4o"]
    fal:
      api_key: "${FAL_KEY}"
      models: ["flux-pro", "stable-diffusion-3"]
    heygen:
      api_key: "${HEYGEN_API_KEY}"
    elevenlabs:
      api_key: "${ELEVENLABS_API_KEY}"
    runway:
      api_key: "${RUNWAY_API_KEY}"
    suno:
      api_key: "${SUNO_API_KEY}"
  defaults:
    image_model: "dall-e-3"
    image_size: "1024x1024"
    tts_voice: "alloy"
  cost_tracking: true
  max_concurrent_jobs: 3

Supported Providers¶

Category	Providers
Image	OpenAI (DALL-E), FAL (Flux/SD), Replicate, Midjourney API
Video	Runway Gen-3, Pika, Kling, HeyGen
Music	Suno, Udio, Stable Audio
TTS	ElevenLabs, OpenAI TTS, Play.ht
LLM	OpenAI, Anthropic, Google, DeepSeek, Ollama

Key Workflows¶

Generate an image¶

Generate a 1024x1024 image of "a futuristic AI operations dashboard" using Flux Pro.

Create a video¶

Create a 10-second video clip showing a data pipeline animation using Runway.

Generate a voiceover¶

Convert this script to audio using ElevenLabs with voice "Adam":
"CorpusIQ helps business operators automate their workflows with AI agents."

Check job status¶

What's the status of my image generation job?
Show me my RunAPI cost breakdown for this month.

Verification¶

# List available providers
hermes chat -q "List all configured RunAPI providers and their status"

# Test image generation
hermes chat -q "Generate a simple test image using the cheapest provider"

Pitfalls¶

⚠️ API costs: AI media APIs are expensive. A single video generation can cost $1-5. Set cost_tracking: true and monitor usage.
⚠️ API key sprawl: This skill requires many API keys. Use a .env file or secrets manager - never hardcode keys in config.
Async jobs: Video and music generation are async (minutes to hours). Use the job queue to check status. Don't block the session waiting.
Rate limits: Free-tier API keys have severe rate limits. For production use, upgrade to paid tiers.
Model availability: Some models (Kling, Midjourney API) have geographic restrictions or waitlists. Test availability before relying on them.
Output storage: Generated media is stored locally. Large outputs can fill disk space. Set an output size cap and auto-cleanup policy.

Integration with CorpusIQ¶

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects