fal.ai Skills

Skills for building inference.sh apps that wrap fal.ai model endpoints.

Authentication

fal.ai API Key: ../.fal.key

Load it before using fal.ai API:

bash

export FAL_KEY=$(cat ../.fal.key)

Workflow: Adding a New fal.ai Model

Use these skills in order:

1. Search for the model

Skill: fal-model-search

Find the fal.ai endpoint and get its OpenAPI schema:

bash

curl "https://api.fal.ai/v1/models?endpoint_id=fal-ai/MODEL_NAME&expand=openapi-3.0"

2. Check if already implemented

Skill: fal-implemented-models

Check ../IMPLEMENTED_MODELS.md to see if model exists.

3. Add the model

Skill: fal-add-model

IMPORTANT: This skill has a strict checklist. Follow each step IN ORDER:

•Fetch OpenAPI schema
•Write MODEL.md - STOP until complete
•Fetch pricing
•Write PRICING.md - STOP until complete
•Run infsh app init
•Implement inference.py
•Test with infsh run
•Update IMPLEMENTED_MODELS.md
•Deploy with infsh deploy

4. Configure pricing

Skill: fal-pricing

Fetch pricing and create CEL expressions for billing.

General inference.sh Skills

Skill: creating-apps

For general inference.sh development not specific to fal.ai:

Skill	Purpose
using-the-cli	CLI commands
writing-app-logic	inference.py patterns
configuring-resources	inf.yml setup
managing-secrets	API keys and secrets
tracking-usage	Output metadata for billing
handling-cancellation	Long-running tasks
optimizing-performance	Best practices
debugging-issues	Troubleshooting

Quick Reference

Task	Skill
"Add fal.ai model X"	fal-add-model
"Find fal.ai models for Y"	fal-model-search
"Is model X implemented?"	fal-implemented-models
"Set up pricing for X"	fal-pricing
"How do I write inference.py?"	creating-apps/writing-app-logic
"Configure GPU resources"	creating-apps/configuring-resources

Directory Structure

code

../
├── .fal.key                    # fal.ai API key
├── IMPLEMENTED_MODELS.md       # Tracking file
├── fal_helper.py               # Shared helper (copy to each app)
├── _skills/                    # These skills
│   ├── SKILL.md                # This file
│   ├── fal-add-model/
│   ├── fal-model-search/
│   ├── fal-implemented-models/
│   ├── fal-pricing/
│   └── creating-apps/
└── [app-name]/                 # Each implemented app
    ├── inference.py
    ├── inf.yml
    ├── requirements.txt
    ├── __init__.py
    ├── fal_helper.py
    ├── MODEL.md                # Schema documentation
    └── PRICING.md              # Pricing documentation