agentleFS
GoogleKnown publishergoogle/capsem/skills/dev-testing-python/SKILL.md

dev-testing-python

Skill179 starsChanged 12 days agoApache-2.0View on GitHub

Python test infrastructure for capsem-builder. Use when running Python tests, checking coverage, or working with golden fixtures and generated schemas.

The file

We split this skill into the rules it gives an agent

2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Import: direct build_system/builder/ mapping to capsem_builder

    line 27
  2. Distinctive:

    If you change the settings schema (node types, metadata fields), all three must be updated together.

    line 69

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Python Testing (capsem-builder)

Quick reference

python3 build_system/scripts/ci/run-bounded-command.py --timeout-seconds 120 -- \
  uv run --project build_system --frozen python -m pytest \
  -c build_system/pyproject.toml --rootdir . build_system/tests/image/ --no-cov

python3 build_system/scripts/ci/run-bounded-command.py --timeout-seconds 60 -- \
  uv run --project build_system --frozen python -m pytest \
  -c build_system/pyproject.toml --rootdir . \
  build_system/tests/image/test_validate.py -k E001 --no-cov

just test <source-commit>  # Complete qualification, governed by AGENTS.md

Package config

build_system/pyproject.toml:

  • Distribution: capsem-builder
  • Import: direct build_system/builder/ mapping to capsem_builder
  • Entry point: capsem-builder = capsem_builder.gatelaunch:builder_main, which selects the project interpreter before dispatching to the image CLI
  • Build: setuptools, with no nested src/ or compatibility package tree
  • Tests: build_system/tests/, using the locked build_system/uv.lock

The gate already lives in build_system/builder/gate/. The installed package mapping and pytest's configured search paths own imports; do not add an obsolete source root to PYTHONPATH. Explicit -c and --rootdir keep mixed image, gate and cross-system tests under the same configuration and import namespace.

Test directory: build_system/tests/image/

FileWhat it covers
test_validate.pyTOML config linting, error codes E001-E305, warnings W001-W012
test_models.pyPydantic image and image-workspace models
test_cli.pyBackend-only Click CLI surface
test_docker.pyJinja rendering and image-build execution primitives
test_manifest.pyBOM collection, manifest rendering, package parsers
test_config.pyTOML loading, defaults generation, roundtrip
test_doctor.pyBuild prerequisite and source-completeness checks
test_audit.pyTrivy/grype parsing and severity summaries
test_image_build_backend.pyPrivate capsem-admin backend command
test_image_module_boundary.pyExact source, package, and entrypoint ownership

Coverage

  • Floors, combined coverage and report locations belong to config/gate.toml and the owning gate steps. A focused --no-cov run is diagnostic feedback.
  • codecov.yml: builder component includes build_system/builder/**

Golden fixtures and cross-language conformance

Golden fixture at tests/settings_spec/golden.json with expected output at tests/settings_spec/expected.json. Three language parsers must produce identical results:

LanguageTest file
Pythontests/test_settings_spec.py
Rustcrates/capsem-core/tests/settings_spec.rs
TypeScriptweb/app/src/lib/__tests__/settings_spec.test.ts

If you change the settings schema (node types, metadata fields), all three must be updated together.

Schema generation pipeline

config/settings/settings.toml -> Pydantic models -> config/settings/schema.generated.json (JSON Schema)
                                                   -> config/settings/ui-metadata.generated.json (UI metadata)
  • just _generate-settings runs the settings-generation owner. Wrap this direct development command with the bounded runner specified in AGENTS.md.
  • Rust reads config/settings/ui-metadata.generated.json via include_str!() in registry.rs
  • TypeScript validates against config/settings/schema.generated.json in conformance tests

In-VM tests (NOT pytest on host)

guest/artifacts/diagnostics/ contains the pytest tests that run INSIDE the VM via just exec "capsem-doctor". These are NOT part of the host uv run --project build_system --frozen pytest suite. They test the guest environment (mounts, networking, sandbox, MCP, runtimes). See /dev-testing-vm for details.

Source layout

build_system/
  pyproject.toml
  uv.lock
  builder/                 # installs as capsem_builder
    image/
        __init__.py
        cli.py           Backend-only Click CLI entry point
        config.py         TOML config loading, defaults generation
        models.py         Pydantic models (GuestImageConfig, ArchConfig, etc.)
        schema.py         Settings schema (SettingsRoot, GroupNode, SettingNode)
        docker.py         Jinja Dockerfile rendering, Docker build execution
        manifest.py       BOM collection, manifest rendering
        validate.py       Compiler-style linting with error codes
        audit.py          Trivy/grype output parsing
        image_build_backend.py Private capsem-admin image build backend
        doctor.py         Build environment doctor checks
  tests/
    image/

Runtime image templates remain under config/docker/; they are not Python package data and do not move into build_system/.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.