neohaskell/NeoHaskell / .agents/skills/neohaskell-test
neohaskell-test
VerifiedRun NeoHaskell doctest examples, all local Cabal test targets, or the maintained Hurl testbed when a framework change needs runtime verification.
NeoHaskell test runs
Use the owned scripts for broad runtime checks. Launch them from the repository root through nix develop --command, as required by Nix Haskell tools. They keep transcripts under .agents-state/neohaskell-test/ and report unavailable checks as SKIP. Add --strict when every discovered target must run. Each child command has a timeout and an 8 MiB output limit.
- Run source examples with
nix develop --command .agents/skills/neohaskell-test/scripts/run-doctest. The script discovers local packages throughcabal.projectand their source settings through each Cabal file. - Before a full service-suite run, complete the PostgreSQL setup. Run every Cabal test-suite component with
nix develop --command .agents/skills/neohaskell-test/scripts/run-full-suite. Cabal receives thealltarget, and the script checks results against test-suite components declared by the project. It then runs the maintained Hurl testbed when Hurl and PostgreSQL are ready. - Run only the maintained Hurl testbed with
nix develop --command .agents/skills/neohaskell-test/scripts/run-testbed --strict. PostgreSQL must already be available at the runner's default local port 5432. A host PostgreSQL service or an isolated Docker PostgreSQL container can supply it. Check that the port is free before starting a container, record its ID, and stop only that container afterward. The owned entrypoint delegates application startup and teardown totestbed/scripts/run-tests.sh. It does not install Hurl or start PostgreSQL. - Run a focused Hspec selection with the TDD runner through
nix develop --command .agents/skills/tdd/scripts/test-match <pattern> [suite]. It rejects empty selections, setup errors, failed assertions, and missing or ambiguous summaries.
Prepare PostgreSQL for service suites
Use an owned disposable PostgreSQL instance at localhost:5432 for the full
suite and full guide evidence runs. Create the neohaskell database and a
neohaskell role with password neohaskell and permission to create databases
when you provision that instance. Other service specs use that database directly.
FILE_UPLOAD_TEST_DB isolates only the file-state fixture.
Before each run, create a separate disposable file-state database and export its
name for the whole command. The fixture resets its tables and rejects an unset
name or neohaskell. Run from the repository root in the same shell:
unset PGOPTIONS
export FILE_UPLOAD_TEST_DB="nh_file_upload_$(python3 -c 'import uuid; print(uuid.uuid4().hex[:12])')"
PGPASSWORD=neohaskell nix develop --command createdb -h localhost -p 5432 -U neohaskell "$FILE_UPLOAD_TEST_DB"
nix develop --command .agents/skills/neohaskell-test/scripts/run-full-suite --strict
Keep PGOPTIONS unset throughout the run. A global fixture schema override can
break the fresh databases that the testbed creates. For full guide evidence,
keep FILE_UPLOAD_TEST_DB exported and run the consumer skill's
agent-docs-check --verify-evidence command after this database setup.
For only the focused file-state fixture, a dedicated disposable database on an existing server is sufficient. Use the same creation command and connection settings above. Run the focused command in the testbed instructions.
These commands exercise the local development build. Use optimized-build, generated-consumer, cross-platform, or CI checks when those contracts matter. The owned scripts run no lint or formatter checks.
The non-strict testbed SKIP result is incomplete coverage. Keep a required PostgreSQL check pending until it passes on the host or Linux Actions at the PR revision. Follow the verification environment decision. Docker supplies only PostgreSQL here, not the framework build.
Framework export, TH marker and generated-API changes also require NeoHaskell consumer verification. Review its starter and testbed migration requirements and run the generated-project contract alongside the affected testbed checks. A passing framework suite does not replace those consumer checks.
For workflow path selection and draft full-CI opt-in, see CI verification.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 319k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 197k | Updated | Claude CodeCodex | Verified | Install |