ras-commander: Skill for Claude Code

eBFE / BLE S3 Catalog Crawl Skill

Crawl FEMA's public BLE S3 bucket and build a local catalog of available Base Level Engineering datasets. Cache results so repeated sessions don't re-crawl.

When to Use

• "What BLE models are available for [state/watershed]?"
• "Get the eBFE catalog for Texas"
• Before any BLE model download or eBFE workflow
• When RasCatalog / RasSources needs to be populated

S3 Source

Bucket: s3://fim-public-availability-data/ (FEMA public, no auth required)

Structure:

fim-public-availability-data/
└── ble/
    └── {state}/
        └── {huc8}/
            ├── Hydraulics/
            │   └── {project}.zip   ← HEC-RAS model
            ├── Terrain/
            └── Mapping/

Alternative bucket (check both):

s3://ras2fim-dev/
s3://fim-dev-outputs/

Crawl Workflow

Step 1: Check Local Cache First

from pathlib import Path
import json

cache_path = Path(".claude/outputs/ble-catalog-cache.json")
if cache_path.exists():
    import time
    age_days = (time.time() - cache_path.stat().st_mtime) / 86400
    if age_days < 30:
        with open(cache_path) as f:
            catalog = json.load(f)
        print(f"Using cached catalog ({age_days:.0f} days old, {len(catalog)} entries)")
        # Skip crawl, use catalog

Step 2: Crawl S3 (if cache miss or stale)

import boto3
from botocore import UNSIGNED
from botocore.config import Config

# Public bucket — no credentials needed
s3 = boto3.client('s3', config=Config(signature_version=UNSIGNED))

bucket = "fim-public-availability-data"
prefix = "ble/"

# Optional: filter by state
state_filter = "TX"  # or None for all states
if state_filter:
    prefix = f"ble/{state_filter}/"

paginator = s3.get_paginator('list_objects_v2')
catalog = []

for page in paginator.paginate(Bucket=bucket, Prefix=prefix, Delimiter='/'):
    for obj in page.get('Contents', []):
        key = obj['Key']
        size = obj['Size']
        # Filter to hydraulics .zip files
        if 'Hydraulics' in key and key.endswith('.zip'):
            parts = key.split('/')
            # parts: ['ble', state, huc8, 'Hydraulics', filename]
            if len(parts) >= 5:
                catalog.append({
                    'state': parts[1],
                    'huc8': parts[2],
                    'filename': parts[-1],
                    's3_path': f"s3://{bucket}/{key}",
                    'size_mb': round(size / 1e6, 1),
                })

print(f"Found {len(catalog)} BLE hydraulic models")