Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Understand and handle CodeRabbit and GitHub API rate limits for review automation. Use when hitting rate limits on @coderabbitai commands, automating review queries, or building scripts that interact with CodeRabbit via the GitHub API. Trigger with phrases like "coderabbit rate limit", "coderabbit throttling", "coderabbit too many requests", "github api rate limit coderabbit".
.claude/skills/jeremylongshore-coderabbit-rate-limits/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 39% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 80% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 61% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 106% | 0% |
Use current dashboard and responses instead of stale concurrency tables. Separate CodeRabbit allowances, file limits, credits, and provider API limits.
references/official-docs.md and re-check any time-sensitive contract before execution.@coderabbitai rate limit reports PR quota without a review.Treat Git-provider sessions, CodeRabbit web sessions, CLI credentials, and CodeRabbit API keys as separate credentials. Use only an already-approved session or secret-manager reference, never print a secret, and do not place credentials in .coderabbit.yaml, source files, logs, or deliverables.
Require billing approval before enabling usage billing or buying credits; never spam retries. Keep analysis and drafts local until approval is explicit, and record who approved the action and its scope.
A limit record, classification, current contract, mitigation, spend impact, and verification. Include source dates, unknowns, and the exact boundary between observed fact and recommendation.
| Condition | Response | |---|---| | Current contract is unclear or docs disagree | Stop mutation, cite both sources, and request owner resolution. | | Required access or approval is missing | Produce a draft and evidence plan only. | | Validation or pilot behavior differs from expectation | Restore the prior state and retain the failed evidence. | | Output contains secrets or private code | Stop, quarantine the artifact, redact it, and notify the data owner. |
Use the rate-limit command without consuming a review.
Separate a provider 403 from CLI allowance response.
references/official-docs.md.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 12,706 | 10,897 | -14% | 1 | 1 | 0% | 2,127 | 4,003 | +88% | 0 | 0 | — |
case-15 | fail→fail | 13,033 | 8,443 | -35% | 1 | 1 | 0% | 2,235 | 3,670 | +64% | 0 | 0 | — |
case-01 | fail→fail | 21,627 | 12,186 | -44% | 1 | 1 | 0% | 4,729 | 4,529 | -4% | 0 | 0 | — |
case-02 | fail→fail | 21,866 | 17,987 | -18% | 1 | 1 | 0% | 4,461 | 5,712 | +28% | 0 | 0 | — |
case-03 | pass→pass | 11,886 | 10,871 | -9% | 1 | 1 | 0% | 2,230 | 4,245 | +90% | 0 | 0 | — |
case-04 | pass→pass | 8,022 | 6,683 | -17% | 1 | 1 | 0% | 1,454 | 3,241 | +123% | 0 | 0 | — |
case-06 | fail→pass | 10,305 | 1,796 | -83% | 1 | 1 | 0% | 1,710 | 2,382 | +39% | 0 | 0 | — |
case-07 | fail→pass | 13,576 | 9,917 | -27% | 1 | 1 | 0% | 2,180 | 3,928 | +80% | 0 | 0 | — |
case-08 | pass→pass | 7,874 | 4,380 | -44% | 1 | 1 | 0% | 1,538 | 2,989 | +94% | 0 | 0 | — |
case-09 | fail→pass | 22,964 | 3,405 | -85% | 1 | 1 | 0% | 2,168 | 2,798 | +29% | 0 | 0 | — |
case-10 | pass→pass | 11,260 | 3,899 | -65% | 1 | 1 | 0% | 2,114 | 2,747 | +30% | 0 | 0 | — |
case-11 | fail→fail | 10,013 | 8,511 | -15% | 1 | 1 | 0% | 1,691 | 3,552 | +110% | 0 | 0 | — |
case-12 | fail→pass | 10,458 | 3,676 | -65% | 1 | 1 | 0% | 1,682 | 2,715 | +61% | 0 | 0 | — |
case-13 | pass→pass | 6,619 | 2,884 | -56% | 1 | 1 | 0% | 1,037 | 2,585 | +149% | 0 | 0 | — |
case-14 | fail→fail | 9,323 | 7,086 | -24% | 1 | 1 | 0% | 2,031 | 3,575 | +76% | 0 | 0 | — |
case-16 | pass→pass | 10,356 | 4,413 | -57% | 1 | 1 | 0% | 2,028 | 2,947 | +45% | 0 | 0 | — |
case-17 | fail→fail | 12,078 | 8,934 | -26% | 1 | 1 | 0% | 2,265 | 3,670 | +62% | 0 | 0 | — |
case-18 | pass→pass | 10,215 | 7,393 | -28% | 1 | 1 | 0% | 1,755 | 3,394 | +93% | 0 | 0 | — |
case-19 | fail→pass | 9,311 | 6,631 | -29% | 1 | 1 | 0% | 1,546 | 3,185 | +106% | 0 | 0 | — |
case-20 | fail→pass | 12,069 | 8,444 | -30% | 1 | 1 | 0% | 2,027 | 3,647 | +80% | 0 | 0 | — |
case-21 | pass→pass | 10,046 | 9,400 | -6% | 1 | 1 | 0% | 1,854 | 3,951 | +113% | 0 | 0 | — |
case-22 | fail→pass | 9,774 | 3,536 | -64% | 1 | 1 | 0% | 1,638 | 2,710 | +65% | 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 +32 percentage points is the difference between those two pass rates over the 22 comparable cases.
The publisher has shipped newer versions since this run, so these numbers describe v1, not the version currently listed.
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.