Contributing to InferCost

Thank you for your interest in contributing to InferCost! This project makes on-premises AI inference costs visible, attributable, and actionable.

Table of Contents

• Code of Conduct
• How Can I Contribute?
• Development Setup
• Making Changes
• Pull Request Process
• Coding Standards
• Community

Code of Conduct

Please read our Code of Conduct. We are committed to providing a welcoming and inclusive environment for everyone.

How Can I Contribute?

Reporting Bugs

Before creating bug reports, please check existing issues.

Good bug reports include:

• Clear, descriptive title
• Steps to reproduce the problem
• Expected vs actual behavior
• InferCost version, Kubernetes version, GPU type
• CostProfile YAML and controller logs

Use the bug report template when creating issues.

Suggesting Features

We track feature requests via GitHub Issues with the enhancement label.

Good feature requests include:

• Clear use case and problem statement
• Proposed solution (if you have one)
• Impact on existing functionality

First-Time Contributors

Look for issues labeled:

• good-first-issue — Small, well-defined tasks
• help-wanted — Larger tasks where we need help
• documentation — Documentation improvements

Areas We Need Help

High Priority:

• Additional GPU CostProfile examples (A100, L40S, RTX 4090, etc.)
• vLLM metrics scraper support
• Helm chart
• Cloud pricing updates and verification

Medium Priority:

• Multi-cluster aggregation
• LiteLLM PostgreSQL integration for per-user attribution
• TokenBudget CRD with PrometheusRule generation
• FOCUS spec export format

Development Setup

Prerequisites

• Go 1.26+: Install from golang.org
• Docker: For building container images
• kubectl: Configured with a Kubernetes cluster
• Kubebuilder: brew install kubebuilder (or install manually)

Clone and Build

git clone git@github.com:defilantech/infercost.git
cd infercost

go mod download
make manifests
make build          # Builds controller + CLI
make test           # Run tests
make lint           # Run linter

Running Locally

# Install CRDs into your cluster
make install

# Run the controller locally
go run ./cmd/main.go \
  --metrics-bind-address=:8090 \
  --metrics-secure=false \
  --health-probe-bind-address=:8091 \
  --dcgm-endpoint=http://<dcgm-exporter>:9400/metrics

# Apply a CostProfile
kubectl apply -f config/samples/finops_v1alpha1_costprofile.yaml

# Check results
kubectl get costprofiles

Making Changes

Branching Strategy

• main — Stable, production-ready code
• feat/* — New features
• fix/* — Bug fixes
• docs/* — Documentation changes

Commit Messages

We use descriptive commit messages with conventional prefixes for Release Please:

Prefix	When to use	Version Bump
feat:	New features, CRD fields	Minor (0.x.0)
fix:	Bug fixes	Patch (0.0.x)
docs:	Documentation only	Patch
chore:	CI, deps, tooling	None
test:	Test-only changes	None

All commits must be signed off (git commit -s) per the Developer Certificate of Origin.

Testing

make test            # Unit tests (envtest)
make lint            # golangci-lint
make manifests       # Verify CRDs are up to date
git diff --exit-code # Should show no changes

Writing tests:

• Table-driven tests for multiple cases
• Test both success and error paths
• Use httptest for HTTP-dependent tests (scraper, API)
• Test CRD validation
• Verify Prometheus metrics are correctly set

Pull Request Process

Before Submitting

• make test passes
• make lint passes
• All commits are signed off (git commit -s)
• Documentation updated (if user-facing change)
• Branch is up-to-date with main

PR Title Format

PR titles drive changelog generation via Release Please:

feat: Add vLLM metrics scraper support
fix: Correct cloud comparison when token count is zero
docs: Add H100 CostProfile example

Review Process

1. Automated checks run (tests, lint, DCO)
2. Maintainer review (usually within 2-3 days)
3. Address feedback by pushing new commits
4. Approval and squash-merge to main

Coding Standards

Go Code

• Follow Effective Go
• Use gofmt and golangci-lint
• Keep functions focused and testable
• Add comments for exported types/functions

CRD Design

• Follow Kubernetes API conventions
• Use +kubebuilder markers for validation
• Provide meaningful status conditions
• Add examples in config/samples/

Cost Calculations

• Document the formula being implemented
• Include units in variable names (e.g., powerDrawWatts, costPerHourUSD)
• Use float64 for all monetary values
• Always show your math in code comments for non-obvious calculations

Community

• GitHub Issues: Bug reports, feature requests
• GitHub Discussions: Q&A, ideas, general discussion

License

By contributing to InferCost, you agree that your contributions will be licensed under the Apache License 2.0.

---

Thank you for contributing to InferCost! Every PR, issue report, and doc improvement helps make AI inference cost tracking accessible to everyone.