[elhoim]

BLUF — Exhausted Qintel API retries return None instead of raising, leaving callers an opaque error.

• Problem — The retry loop in _search (misp_modules/lib/qintel_helper.py) guards its raise with an inner if request_attempts < max_retries that duplicates the while condition and is evaluated before the counter increments, so the else branch is dead and _search falls off the end returning None.
• Fix — Remove the redundant condition, dedent the wait/log/sleep block, and move raise Exception("Max API retries exceeded") after the loop.
• Effect — When the Qintel API rate-limits or times out repeatedly, the five callers get that informative exception instead of an opaque AttributeError on NoneType.read().

The retry loop in qintel_helper.py's _search had a dead error branch:

while request_attempts < max_retries:
    ...
    if request_attempts < max_retries:
        wait_time = _get_request_wait_time(request_attempts)
        ...
        sleep(wait_time)
    else:
        raise Exception("Max API retries exceeded")

    request_attempts += 1

The inner if request_attempts < max_retries duplicates the while loop's own guard and is evaluated before request_attempts is incremented, so it is always true inside the loop body — the else branch (and its raise) can never execute. When retries are actually exhausted, the while condition simply becomes false and _search falls off the end of the function, returning None instead of raising.

Impact: All five callers of _search (search_pmi, search_qwatch, search_qauth, search_qsentry, qsentry_feed) immediately call .read() on the result. When the Qintel API rate-limits or times out repeatedly, the analyst doesn't see the intended Exception("Max API retries exceeded") — they get an opaque AttributeError: 'NoneType' object has no attribute 'read' deep in the module, with no indication that the real cause was exhausted retries against the Qintel API.

Fix: Removed the redundant inner condition, dedented the wait/log/sleep block so it always runs inside the loop, and moved raise Exception("Max API retries exceeded") to after the while loop, so it fires exactly when the loop exits because retries are exhausted. This is the only behavior change: on retry exhaustion the caller now gets the intended, informative exception instead of an unrelated AttributeError several frames later.

Verification

• python -m py_compile misp_modules/lib/qintel_helper.py — clean. (This file lives under misp_modules/lib/, which the repo's flake8 CI config excludes.)
• Full module test suite: 161 passed, 4 skipped, 5 subtests passed in 176.98s (0:02:56), matching the expected baseline exactly.

Found during a review of the repository; other findings are being submitted as separate PRs.

🤖 Generated with Claude Code