mcpField
F
finops-compare

finops-compare

github.com/laurigatesCloud & DevOps5954 installsUpdated 1d ago
Repository

About finops-compare

A skill for comparing GitHub Actions FinOps metrics across repositories in a GitHub organization. It helps teams identify differences in cache use, workflow activity, failures, and efficiency.

Key features

  1. 01Compare cache usage across repositories
  2. 02Compare workflow frequency and failures
  3. 03Discover organization repositories
  4. 04Limit discovery or specify repositories

Use cases

  1. 01Find repositories with high cache usage
  2. 02Audit workflow failures across an organization
  3. 03Compare CI metrics for selected repositories

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/laurigates/claude-plugins --skill finops-compare

After installing, type/finops-comparein Claude Code to confirm it loaded.

Bundled filesscripts/compare-repos.sh

Details

From the project README

/finops:compare

Compare GitHub Actions FinOps metrics across multiple repositories - cache usage, workflow frequency, failure rates, and efficiency.

When to Use

ScenarioUse this skillAlternative
Compare CI metrics across org repos/finops:compare-
Identify worst-performing repos in org/finops:compare-
Org-wide cache or failure audit/finops:compare-
Deep-dive into a single repo's caches/finops:cachesUse caches for single-repo detail
Deep-dive into a single repo's workflows/finops:workflowsUse workflows for per-repo analysis
Quick summary of current repo only/finops:overviewUse overview for single-repo snapshot

Parameters

ParameterDescriptionDefault
orgGitHub organization name (required)-
repos...Space-separated list of repo namesAll org repos
--limit NLimit auto-discovery to N repos30

Usage Examples

# Compare specific repos
/finops:compare myorg repo1 repo2 repo3

# Compare all repos in org (up to 30)
/finops:compare myorg

# Compare more repos
/finops:compare myorg --limit 50

Execution

bash "${SKILL_DIR}/scripts/compare-repos.sh" $ARGS

Output Format

=== FinOps Comparison: myorg ===

Discovering repos (limit: 30)...
Found 25 repos

=== Cache Usage ===
Repository                               Caches    Size (MB)
----------                               ------    ---------
frontend-app                                 45         2340
backend-api                                  32         1850
shared-libs                                  18          420
...

=== Workflow Activity (last 30 days) ===
Repository                                 Runs  Success   Failed  Skip Rate
----------                                 ----  -------   ------  ---------
frontend-app                                156      140       10         3%
backend-api                                  89       85        2         2%
...

=== Failure Rates (top 15) ===
Repository                                Total   Failed       Rate
----------                                -----   ------       ----
legacy-service                               45       12        26%
experimental-repo                            20        5        25%
...

=== Active Workflows ===
Repository                               Workflows
----------                               ---------
frontend-app                                     8
backend-api                                      5
...

=== Summary ===
Total cache usage: 8450MB across 25 repos

Repos exceeding 1GB cache:
  frontend-app: 2340MB
  backend-api: 1850MB

Repos with >20% failure rate:
  legacy-service: 26%
  experimental-repo: 25%

Agentic Optimizations

ContextCommand
List org repos (JSON)gh api "/orgs/{org}/repos?per_page=100&sort=pushed" --jq '.[].full_name'
Cache count per repogh api "/repos/{owner}/{repo}/actions/caches" --jq '.total_count'
Workflow runs (JSON)gh api "/repos/{owner}/{repo}/actions/runs?per_page=100" --jq '.workflow_runs'
Compact multi-repo comparebash "${SKILL_DIR}/scripts/compare-repos.sh" $ARGS

Post-actions

Based on comparison results:

  • High cache repos: Run /finops:caches <repo> for detailed analysis
  • High failure repos: Run /finops:workflows <repo> to investigate
  • High activity repos: Run /finops:waste <repo> to find optimizations
  • Create report: Consider creating a GitHub issue with findings

Comments

Markdown supported