Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Check for new OpenClaw releases and notify once per new version.
.claude/skills/jdrhyne-clawdbot-release-check/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -55% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -39% | 0% |
Checks for new OpenClaw releases from GitHub and notifies once per version (no spam).
bashclawhub install clawdbot-release-check
bash# Add daily update check at 9am, notify via Telegram {baseDir}/scripts/setup.sh --telegram YOUR_TELEGRAM_ID # Custom hour {baseDir}/scripts/setup.sh --hour 8 --telegram YOUR_TELEGRAM_ID # Remove cron job {baseDir}/scripts/setup.sh --uninstall
After setup, restart gateway:
bashopenclaw gateway restart
bash{baseDir}/scripts/check.sh {baseDir}/scripts/check.sh --status {baseDir}/scripts/check.sh --force {baseDir}/scripts/check.sh --all-highlights {baseDir}/scripts/check.sh --reset {baseDir}/scripts/check.sh --help
~/.openclaw/clawdbot-release-check-state.json~/.openclaw/clawdbot-release-check-cache.jsonEnvironment variables:
OPENCLAW_DIR — path to OpenClaw source/installCACHE_MAX_AGE_HOURS — cache TTL in hours (default: 24)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 10,263 | 3,834 | -63% | 1 | 1 | 0% | 1,700 | 942 | -45% | 0 | 0 | — |
case-02 | fail→pass | 10,217 | 3,509 | -66% | 1 | 1 | 0% | 1,666 | 889 | -47% | 0 | 0 | — |
case-03 | fail→pass | 8,276 | 4,331 | -48% | 1 | 1 | 0% | 1,423 | 1,058 | -26% | 0 | 0 | — |
case-04 | fail→pass | 9,506 | 2,187 | -77% | 1 | 1 | 0% | 1,510 | 673 | -55% | 0 | 0 | — |
case-05 | fail→pass | 19,654 | 2,360 | -88% | 1 | 1 | 0% | 1,168 | 713 | -39% | 0 | 0 | — |
case-06 | fail→pass | 8,698 | 2,296 | -74% | 1 | 1 | 0% | 1,423 | 667 | -53% | 0 | 0 | — |
case-07 | pass→pass | 10,092 | 1,804 | -82% | 1 | 1 | 0% | 1,765 | 598 | -66% | 0 | 0 | — |
case-08 | fail→pass | 21,579 | 1,953 | -91% | 1 | 1 | 0% | 1,600 | 596 | -63% | 0 | 0 | — |
case-09 | fail→pass | 8,785 | 1,817 | -79% | 1 | 1 | 0% | 1,436 | 566 | -61% | 0 | 0 | — |
case-10 | fail→pass | 6,170 | 4,287 | -31% | 1 | 1 | 0% | 915 | 710 | -22% | 0 | 0 | — |
case-11 | fail→pass | 9,500 | 2,014 | -79% | 1 | 1 | 0% | 1,488 | 640 | -57% | 0 | 0 | — |
case-12 | pass→pass | 3,303 | 1,657 | -50% | 1 | 1 | 0% | 486 | 511 | +5% | 0 | 0 | — |
case-13 | pass→pass | 8,803 | 2,126 | -76% | 1 | 1 | 0% | 1,385 | 591 | -57% | 0 | 0 | — |
case-14 | pass→pass | 9,745 | 4,346 | -55% | 1 | 1 | 0% | 1,704 | 1,058 | -38% | 0 | 0 | — |
case-15 | pass→pass | 6,063 | 3,457 | -43% | 1 | 1 | 0% | 1,086 | 897 | -17% | 0 | 0 | — |
case-16 | pass→pass | 2,924 | 2,428 | -17% | 1 | 1 | 0% | 472 | 683 | +45% | 0 | 0 | — |
case-17 | fail→pass | 8,444 | 1,927 | -77% | 1 | 1 | 0% | 1,336 | 621 | -54% | 0 | 0 | — |
case-18 | fail→pass | 10,382 | 2,567 | -75% | 1 | 1 | 0% | 1,774 | 751 | -58% | 0 | 0 | — |
case-19 | pass→pass | 9,223 | 5,346 | -42% | 1 | 1 | 0% | 1,641 | 1,246 | -24% | 0 | 0 | — |
case-20 | pass→pass | 7,508 | 2,127 | -72% | 1 | 1 | 0% | 1,140 | 635 | -44% | 0 | 0 | — |
case-21 | pass→pass | 10,410 | 7,487 | -28% | 1 | 1 | 0% | 1,674 | 1,589 | -5% | 0 | 0 | — |
case-22 | fail→pass | 7,752 | 1,966 | -75% | 1 | 1 | 0% | 1,176 | 589 | -50% | 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. 22 cases were attempted. The headline lift of +59 percentage points is the difference between those two pass rates over the 22 comparable cases.
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.