tenacity exception classifier

I have completed the audit of your retry logic. I identified 3 critical issues where permanent failures were being retried and root causes were being swallowed by Tenacity's default behavior.

Tenacity Exception Audit Report

Detailed Findings & Fixes

Finding EXC-3-4a1b: Dangerous HTTP Retries
The current implementation retries on any httpx.HTTPStatusError. This includes 404 Not Found and 401 Unauthorized, which will never succeed on a retry and may trigger security lockouts.

Recommended Fix:
Use a transient-only predicate. I have generated a patch that implements a status-code filter.

# Generated Patch: ./output/patches/EXC-3-4a1b.py
import httpx
from tenacity import retry, retry_if_exception, stop_after_attempt, wait_exponential_jitter

_TRANSIENT_CODES = {408, 429, 500, 502, 503, 504}

def is_transient_http_error(exc: BaseException) -> bool:
    if isinstance(exc, httpx.ConnectError):
        return True
    if isinstance(exc, httpx.HTTPStatusError):
        return exc.response.status_code in _TRANSIENT_CODES
    return False

@retry(
    retry=retry_if_exception(is_transient_http_error),
    stop=stop_after_attempt(3),
    wait=wait_exponential_jitter(initial=1, max=10),
    reraise=True
)
def fetch_data():
    ...

Next steps

1. Review the full report in ./output/tenacity-exception-classifier-report.md.
2. Apply the specific fixes from ./output/patches/.
3. Verify that your logging captures the is_transient_http_error predicate logic.