EXTENSIONS
Gcp/testing
Google Cloud testing infrastructure models
Docker Compose
Two methods for `@smith/docker-compose` that end-to-end setups need.
Salmon
A throwaway Salmon for end-to-end tests of anything that takes Salmon's tokens: database validators, message brokers, services.
Kind
One named kind (Kubernetes in Docker) cluster for swamp, created on demand for end-to-end tests.
Mere Pkgd Harness
Local integration-test harness for the mere <-> pkgd package-publish pipeline: builds both binaries, runs pkgd as a real background process against scratch directories, publishes a real signed test package to it over HTTP, and verifies the result — without touching the real /mere on the host.
Datadog/synthetics
Datadog Synthetics — synthetic monitoring tests, results, and locations
Test Evidence Validator
Validates a caller-supplied parsed test result for internal consistency and a recognized multi-language test-runner command, then commit-pins the evidence to an immutable Git commit and branch, returning a single pass/fail verdict.
Herdr Swamp Fixtures
Harmless test workflows (a manual approval gate and a slow failing step) for trying out a swamp dashboard such as the Herdr swamp layout.
Unity
Drive the Unity Editor in batch mode with its documented command-line
Swamp Training
Retest the swamp-training Slidev deck when swamp ships a new release: every swamp command in the deck is checked against the release's CLI, the deck's @training/file-check lab is rebuilt from the slides and run, and the skills, CLAUDE.md and allowlist that swamp repo init installs are diffed against the last tested release. The release binary runs in a throwaway sandbox, so your installed swamp is left alone.
Floci Aws
Floci AWS protocol compatibility models and reproducible S3 lifecycle tests for Swamp.
Kind Cluster Pool
Desired-size pool of local kind clusters with reserve, release, reconciliation, and status operations for test infrastructure.
Docker Image Test
Local Docker image smoke testing for swamp: build image matrices, run containers, poll health checks, capture logs, and clean up.