Skillshub coderabbit-cost-tuning
install
source · Clone the upstream repo
git clone https://github.com/ComeOnOliver/skillshub
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/ComeOnOliver/skillshub "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/jeremylongshore/claude-code-plugins-plus-skills/coderabbit-cost-tuning" ~/.claude/skills/comeonoliver-skillshub-coderabbit-cost-tuning && rm -rf "$T"
manifest:
skills/jeremylongshore/claude-code-plugins-plus-skills/coderabbit-cost-tuning/SKILL.mdsource content
CodeRabbit Cost Tuning
Overview
Optimize CodeRabbit per-seat licensing costs by right-sizing seat allocation, focusing reviews on high-value repositories, and configuring review scope to minimize unnecessary AI processing. CodeRabbit charges per seat based on active committers who open PRs.
Prerequisites
- CodeRabbit Pro or Enterprise plan
- GitHub/GitLab org admin access
- Access to CodeRabbit dashboard at app.coderabbit.ai
Pricing Model
| Plan | Price | Includes |
|---|---|---|
| Free | $0 | Public repos, limited reviews |
| Pro | ~$15/seat/month | Unlimited reviews, all features |
| Enterprise | Custom | SSO, dedicated support, SLA |
Seat = any developer who opens a PR in a CodeRabbit-enabled repo.
Instructions
Step 1: Audit Seat Utilization
Navigate to CodeRabbit Dashboard > Organization > Seats:
# Example seat audit seat_audit: active_committers_30d: 15 # These cost money bot_accounts: 3 # Dependabot, Renovate, CI bots (should NOT consume seats) inactive_30d: 7 # Haven't opened a PR in 30 days total_seats_billed: 25 # Savings: Remove bots (3) + inactive (7) = 10 fewer seats # At ~$15/seat/month = $150/month savings
Step 2: Set Seat Policy to Active Committers Only
In CodeRabbit Dashboard > Organization > Billing:
- Switch seat policy from "All org members" to "Active committers"
- Define active as "opened a PR in the last 30 days"
- Exclude bot accounts explicitly:
,dependabot[bot]
,renovate[bot]github-actions[bot]
Step 3: Focus Reviews on High-Value Repos
Only enable CodeRabbit on repos where AI review adds value:
# Enable CodeRabbit (high value): - backend-api → Business logic, security-critical - payment-service → PCI compliance, financial data - infrastructure → Terraform/IaC, blast radius high - mobile-app → Customer-facing, release quality # Disable CodeRabbit (low value): - documentation → Markdown only, low risk - design-assets → Binary files, not reviewable - sandbox → Experimental, throwaway code - archived-* → Read-only repos - internal-tools → Low-traffic, single-developer repos # To disable: GitHub > Installed Apps > CodeRabbit > Repository access # Switch to "Only select repositories" and remove low-value repos
Step 4: Exclude Low-Value Files from Reviews
# .coderabbit.yaml - Skip files that don't benefit from AI review reviews: path_filters: - "!**/*.lock" # Lock files (no actionable feedback) - "!**/package-lock.json" - "!**/pnpm-lock.yaml" - "!**/*.snap" # Test snapshots - "!**/*.generated.*" # Generated code - "!dist/**" # Build output - "!vendor/**" # Third-party code - "!**/*.min.js" # Minified files - "!**/migrations/*.sql" # DB migrations (review manually) - "!**/*.csv" # Data files - "!**/*.json" # Config/data files (usually low-value) auto_review: ignore_title_keywords: - "chore: bump" # Skip dependency update PRs - "chore(deps)" - "auto-generated" - "Bump version" drafts: false # Don't burn credits reviewing drafts
Step 5: Use the Right Review Profile
# More aggressive profile = more comments = more processing # But the main cost is per-seat, not per-comment reviews: profile: "assertive" # Recommended default # "chill" produces fewer comments but same per-seat cost # Choose based on signal-to-noise, not cost optimization
Step 6: Monitor Review Value
Track whether CodeRabbit reviews are being acted on:
set -euo pipefail ORG="${1:-your-org}" REPO="${2:-your-repo}" echo "=== CodeRabbit Review Value Analysis ===" TOTAL_PRS=0 REVIEWED_PRS=0 for PR_NUM in $(gh api "repos/$ORG/$REPO/pulls?state=closed&per_page=30" --jq '.[].number'); do TOTAL_PRS=$((TOTAL_PRS + 1)) CR_COMMENTS=$(gh api "repos/$ORG/$REPO/pulls/$PR_NUM/comments" \ --jq '[.[] | select(.user.login=="coderabbitai[bot]")] | length' 2>/dev/null || echo "0") if [ "$CR_COMMENTS" -gt 0 ]; then REVIEWED_PRS=$((REVIEWED_PRS + 1)) echo "PR #$PR_NUM: $CR_COMMENTS CodeRabbit comments" fi done echo "" echo "Coverage: $REVIEWED_PRS/$TOTAL_PRS PRs received CodeRabbit reviews" echo "If coverage is low, check: base_branches filter, drafts setting, seat assignment"
Step 7: CLI Credit Management
# CodeRabbit CLI charges per file reviewed (~$0.25/file) # Tips to reduce CLI costs: # Review only specific files (not entire repo) cr review src/api/routes.ts src/middleware/auth.ts # Use --prompt-only to get review text without interactive mode cr review --prompt-only # Set up pre-push hook (not pre-commit) to avoid reviewing WIP code # See coderabbit-local-dev-loop for hook setup
Output
- Seat audit completed with wasted seats identified
- Repository access scoped to high-value repos only
- Path filters configured to skip low-value files
- Review coverage metrics measured
- CLI usage optimized with targeted file reviews
Error Handling
| Issue | Cause | Solution |
|---|---|---|
| Seat count higher than expected | Bots counted as seats | Exclude bot accounts in dashboard |
| Reviews on archived repos | App still installed | Remove CodeRabbit from archived repos |
| Low review acceptance rate | Reviews too nitpicky | Switch profile to |
| Can't reduce seat count | Active committers across all repos | Disable CodeRabbit on low-value repos first |
| CLI charges higher than expected | Reviewing all files | Use instead |
Resources
Next Steps
For enterprise seat management and SSO, see
coderabbit-enterprise-rbac.