CommunityResearch & Data Analysisgithub.com

Nikkiwei129/tibo-reset-early-warning

A low-model-usage Agent Skill that monitors public signals for possible future Codex quota resets.

What is tibo-reset-early-warning?

tibo-reset-early-warning is a Codex agent skill that a low-model-usage Agent Skill that monitors public signals for possible future Codex quota resets.

Works with~Claude CodeCodex CLI~Cursor
npx skills add Nikkiwei129/tibo-reset-early-warning

Installed? Explore more Research & Data Analysis skills: obra/superpowers, affaan-m/quarkus-verification, affaan-m/uspto-database · View all 6 →

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

Tibo Reset Early Warning

Provide an early warning before a possible Codex quota reset without waking a model for every routine poll.

Operating model

  1. Run scripts/check_feed.py locally on a schedule. It uses Python's standard library, a public feed, and local state; routine runs do not call a model.
  2. Treat exit code 0 with no output as normal silence. Treat JSON Lines on stdout as new candidates; only then assess the text semantically.
  3. Notify for future signals: red means explicit future timing or rollout; yellow means a credible tease or preparation signal.
  4. Ignore jokes, negations, unrelated resets, API rate limits, and posts that only confirm a completed reset.
  5. State the original post time, link, confidence, and inferred Beijing-time window. Clearly label inference.
  6. Recommend starting valuable queued work only. Never recommend empty token burning.

If the feed is stale, unavailable, malformed, or not scoped to @thsottiaux, send no quota alert. Surface the operational error in logs.

Commands

python3 scripts/check_feed.py --state /tmp/tibo-reset-state.json --dry-run
python3 scripts/check_feed.py --state "$TIBO_RESET_WARNING_STATE"

The first normal run establishes a baseline and emits no historical alerts. Use --replay only for testing.

For scheduling and notification wiring, read references/scheduling.md. For publishing with GitHub Desktop, read references/github-desktop.md.

Related Skills