mirror of
https://github.com/trustedsec/hate_crack.git
synced 2026-04-28 12:03:11 -07:00
Introduce hate_crack.notify package with a small functional public API and a CrackTailer thread for polling hashcat output files. Package layout keeps the HTTP call (_send_pushover) isolated so future backends (Slack, generic webhooks) can be added as a sibling function rather than a framework rewrite. Core pieces: - settings.py: NotifySettings dataclass plus atomic config persistence (save_enabled, add_to_allowlist) via read-modify-write + os.replace. - pushover.py: single _send_pushover() that never raises; network errors, missing requests, and missing creds all funnel to False. - _suppress.py: thread-local suppression context manager so orchestrator attacks can chain primitives without flooding notifications. - tailer.py: CrackTailer(threading.Thread) that seeks to EOF on start, polls at a user-configurable interval, and collapses per-tick bursts into a single aggregate notification when they exceed the cap. - __init__.py: public API (init, prompt_notify_for_attack, notify_job_done, notify_crack, start_tailer, stop_tailer, toggle_enabled, suppressed_notifications). Privacy guarantee: notification payloads contain only attack name, counts, and hash path, never plaintexts. 72 new tests cover dataclass defaults, atomic config writes, idempotent allowlist updates, HTTP payload privacy, suppression nesting and thread-locality, tailer EOF seek, burst cap, truncation recovery, and the per-attack prompt's [y/n/always] flow. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
7.7 KiB
7.7 KiB