| name | api-proxy |
| description | Routes API requests through the user's residential proxy to bypass cloud IP blocks. TWO use cases - (1) Runtime fallback when direct curl/fetch fails with 403/blocked errors, use fetch.py for quick retry. (2) Template for building new skills that need proxy routing, copy proxy_helper.py into the new skill. |
API Proxy Skill
Routes requests through residential IP via proxy.address.here.com.
Runtime Fallback (ad-hoc)
When a direct request fails with 403 or IP blocking, retry through the proxy:
python3 /mnt/skills/user/api-proxy/scripts/fetch.py "https://api.example.com/endpoint"
python3 /mnt/skills/user/api-proxy/scripts/fetch.py "https://api.example.com/endpoint" POST '{"key": "value"}'
Returns JSON response or {"error": "message"} on failure.
Template Mode (building skills)
When creating skills that need proxy routing:
- Copy
proxy_helper.py into the new skill's scripts/ folder
- Import and use:
from proxy_helper import proxy_get, proxy_request
data = proxy_get("https://api.example.com/endpoint")
data = proxy_request(
url="https://api.example.com/endpoint",
method="POST",
headers={"Authorization": "Bearer xyz"},
body='{"key": "value"}'
)
Proxy Details
URL: https://proxy.address.here.com/proxy
Token: your_token_here
Currently Whitelisted Domains
reddit.com, www.reddit.com, oauth.reddit.com
New domains require user to add them to ALLOWED_DOMAINS env var and restart the container.
Troubleshooting
| Error | Cause | Fix |
|---|
| 403 from proxy | Domain not whitelisted | Ask user to whitelist |
| 503 from proxy | Rate limited | Retry after delay |
| 504 from proxy | Upstream timeout | Increase timeout |
| Connection refused | Container/tunnel down | Check status |
