Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Send Bark (day.app) push notifications after Codex completes a task. Use when one Codex or Claude run is finished, or when you need to notify on task completion with device name, project name, status, and summary.
.claude/skills/aiskillstore-bark-notify/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -24% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -46% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -82% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -26% | 0% |
Send a Bark push notification after each Codex task. Use the helper script to resolve the project name and post to the Bark API.
You can add the following lines to ~/.bashrc (Linux) or ~/.zshrc (macOS) to set environment variables:
export CODEX_MACHINE_NAME="Machine-name" # e.g., Macbook
export CODEX_BARK_KEY="" # https://api.day.app/<CODEX_BARK_KEY>/
export CODEX_BARK_BASE_URL="https://api.day.app"For example, you can add this instruction to AGENTS.md: > Use skill bark-notify to notify users when each agent run is finished or when any notifications would be sent to users.
AGENTS.md using one of:project_name: My Project (or name:)Project Name: My Project or 项目名称：我的项目success, failed, partial, blocked, etc.Run:
bashpython3 ~/.codex/skills/bark-notify/scripts/send_bark_notification.py \ --task-title \"...\" \ --status \"success\" \ --summary \"...\" \ --project-name \"...\"
scripts/send_bark_notification.py: Send the Bark notification and resolve the project name.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | fail→pass | 20,056 | 6,983 | -65% | 1 | 1 | 0% | 968 | 732 | -24% | 0 | 0 | — |
case-01 | fail→fail | 4,289 | 11,545 | +169% | 1 | 1 | 0% | 694 | 762 | +10% | 0 | 0 | — |
case-02 | fail→fail | 21,842 | 16,997 | -22% | 1 | 1 | 0% | 1,180 | 720 | -39% | 0 | 0 | — |
case-03 | fail→fail | 6,437 | 12,957 | +101% | 1 | 1 | 0% | 998 | 876 | -12% | 0 | 0 | — |
case-05 | fail→pass | 14,335 | 7,284 | -49% | 1 | 1 | 0% | 1,525 | 829 | -46% | 0 | 0 | — |
case-06 | fail→pass | 29,545 | 7,918 | -73% | 1 | 1 | 0% | 5,317 | 975 | -82% | 0 | 0 | — |
case-07 | fail→pass | 26,665 | 7,269 | -73% | 1 | 1 | 0% | 885 | 741 | -16% | 0 | 0 | — |
case-08 | pass→pass | 12,185 | 7,706 | -37% | 1 | 1 | 0% | 1,189 | 897 | -25% | 0 | 0 | — |
case-09 | fail→pass | 12,397 | 2,557 | -79% | 1 | 1 | 0% | 1,139 | 839 | -26% | 0 | 0 | — |
case-10 | pass→pass | 14,649 | 6,682 | -54% | 1 | 1 | 0% | 1,570 | 692 | -56% | 0 | 0 | — |
case-11 | fail→pass | 10,841 | 2,731 | -75% | 1 | 1 | 0% | 1,947 | 956 | -51% | 0 | 0 | — |
case-12 | fail→pass | 14,349 | 6,465 | -55% | 1 | 1 | 0% | 1,588 | 621 | -61% | 0 | 0 | — |
case-13 | pass→pass | 18,614 | 8,779 | -53% | 1 | 1 | 0% | 1,827 | 907 | -50% | 0 | 0 | — |
case-14 | pass→pass | 13,654 | 7,789 | -43% | 1 | 1 | 0% | 1,382 | 768 | -44% | 0 | 0 | — |
case-15 | fail→pass | 14,695 | 2,603 | -82% | 1 | 1 | 0% | 1,388 | 879 | -37% | 0 | 0 | — |
case-16 | fail→pass | 9,703 | 9,060 | -7% | 1 | 1 | 0% | 712 | 1,157 | +63% | 0 | 0 | — |
case-17 | pass→fail | 12,492 | 16,687 | +34% | 1 | 1 | 0% | 1,240 | 742 | -40% | 0 | 0 | — |
case-18 | pass→pass | 9,359 | 6,753 | -28% | 1 | 1 | 0% | 1,599 | 730 | -54% | 0 | 0 | — |
case-19 | pass→pass | 12,421 | 7,553 | -39% | 1 | 1 | 0% | 1,167 | 855 | -27% | 0 | 0 | — |
case-20 | pass→pass | 17,014 | 15,940 | -6% | 1 | 1 | 0% | 2,442 | 2,535 | +4% | 0 | 0 | — |
case-21 | pass→pass | 18,701 | 9,977 | -47% | 1 | 1 | 0% | 2,043 | 2,452 | +20% | 0 | 0 | — |
case-22 | pass→pass | 11,658 | 10,208 | -12% | 1 | 1 | 0% | 2,151 | 2,466 | +15% | 0 | 0 | — |
case-23 | pass→pass | 13,656 | 6,945 | -49% | 1 | 1 | 0% | 1,583 | 1,817 | +15% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted, and 18 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +35 percentage points is the difference between those two pass rates over the 18 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.