Free CI tools
CI cost and rerun waste calculator
Estimate CI execution cost and rerun waste with your own runner rate, job duration, and extra execution ratio. See assumptions and a worked monthly example.
Use calculatorResources
Practical CI/CD for teams working with agents
Decisions, examples, and evidence for the work between a code change and a verified release.
Free CI tools
Flaky tests: suite and retry probability calculator
Explore how flaky tests compound across a suite and repeated attempts. Calculate illustrative probabilities, with worked examples and clear model assumptions.
Use calculatorFree CI tools
GitHub Actions matrix budget calculator
Estimate matrix jobs, execution waves, elapsed time, and total job minutes. Explore axes and exclusions with assumptions and GitHub Actions limit guidance.
Use calculatorFree CI tools
Test parallelization break-even calculator
Explore when test parallelization outweighs added overhead. Compare modeled duration, speedup, and reserved worker-seconds with explicit limits and examples.
Use calculatorFree CI tools
Deployment frequency calculator
Calculate deployment frequency per day and week from production-release events. See counting rules, a synthetic log, and why implied intervals need context.
Use calculatorCI/CD decisions
Choosing a GitHub Actions alternative for agent-driven teams
Compare runner changes with a full CI/CD migration. A practical guide to agent-operated testing, deployment, and evaluating Agentic Pipeline’s closed beta.
Read the guideTesting agent changes
Testing code written by AI agents
A practical guide to testing AI-generated code against requirements, finding independent counterexamples, and deciding when a generated change is safe to ship.
Read the guideAgent delivery workflows
MCP for CI/CD: a safe build-to-deploy loop
Learn how an agent can use MCP for CI/CD: read exact run evidence, make a bounded repair, follow delivery, and verify the revision that reached production.
Read the guideCI/CD security
GitHub Actions permissions: least privilege by job
Scope GitHub Actions permissions to each job. See a read-only test and write-capable release example, plus OIDC, fork, and token boundaries to review.
Read the guideCI/CD operations
GitHub Actions concurrency and cancellation
Choose GitHub Actions concurrency for stale pull request checks and serialized releases. See pending replacement, queue limits, ordering, and revision proof.
Read the guideCI/CD operations
GitHub Actions cache keys and safe invalidation
Design GitHub Actions cache keys for OS, runtime, and lockfile changes. Learn fallback risks, cache-mode trust boundaries, and why a cache hit proves little.
Read the guideCI/CD governance
Required status checks and branch protection
Learn how GitHub branch protection evaluates required checks by revision, context, and source app, including skipped workflows and merge queue failures.
Read the guideCI/CD operations
GitHub Actions self-hosted runners: isolation and cost
Assess GitHub Actions self-hosted runners by isolation, credential exposure, capacity, and operating cost before moving jobs from hosted compute.
Read the guideCI/CD security
GitHub Actions secrets: boundaries for agent-run workflows
Map when GitHub Actions secrets reach fork tests, reusable workflows, and trusted releases. Separate stored secrets from host credentials and OIDC roles.
Read the guideCI/CD operations
GitHub Actions timeout: bound retries and cleanup
Set GitHub Actions timeouts from observed queue, job, and step delays. Budget bounded retries and cleanup without losing the first failure or deployment proof.
Read the guideCI/CD evidence
GitHub Actions artifacts: useful evidence without false trust
Use GitHub Actions artifacts for test evidence and release handoff while checking origin, digest, retention, and attestation before trusting bytes.
Read the guideCI/CD operations
GitHub Actions workflow_dispatch: safe manual inputs
Design GitHub Actions workflow_dispatch inputs for reviewed refs and targets, then validate them before credentials and verify the resulting run and deployment.
Read the guideCI/CD security
CI/CD security checklist for agent-operated delivery
Review CI/CD security across untrusted pull requests, runners, artifacts, caches, and release identity with concrete gates for agent-operated delivery.
Read the guideDeployment operations
Software deployment checklist: candidate to verified live
Use a software deployment checklist to connect source, build, checks, release operation, and live revision, with stop and recovery steps at each gate.
Read the guideDeployment operations
Deployment rollback: recover code without guessing about data
Choose a deployment rollback from target state and data compatibility. Trace the recovery operation, live revision, and critical journey before closure.
Read the guideDeployment operations
CI/CD monitoring: trace runs to verified deployments
Monitor CI/CD from queue and checks through delivery and live revision. Use correlated evidence, scoped metrics, and alerts for failed or missing proof.
Read the guideDeployment operations
Continuous integration, delivery, and deployment: what each proves
See what continuous integration, delivery, and deployment each prove. Follow a commit through checks, artifact readiness, provider action, and live proof.
Read the guideGitHub Actions
GitHub Actions logs: find the first causal failure
Diagnose GitHub Actions logs by run, attempt, job, and step. Separate the first causal error from cleanup noise, then choose a safe rerun or new commit.
Read the guideCI/CD practice
CI/CD best practices for small teams shipping with coding agents
Build a reliable CI/CD loop for coding agents: small changes, meaningful tests, fast feedback, trusted artifacts, bounded deploys, live proof, and recovery.
Read the guideTesting practice
A practical test pyramid for code written by AI agents
Choose a useful test pyramid for agent-written changes. Map failures to domain, data, contract, and browser boundaries without duplicating fragile checks.
Read the guide