lab-kit-cli 0.6.2__tar.gz → 1.0.0__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- lab_kit_cli-1.0.0/.github/workflows/ci.yml +38 -0
- {lab_kit_cli-0.6.2 → lab_kit_cli-1.0.0}/.github/workflows/release.yml +6 -10
- {lab_kit_cli-0.6.2 → lab_kit_cli-1.0.0}/AGENTS.md +9 -11
- lab_kit_cli-1.0.0/CHANGELOG.md +159 -0
- lab_kit_cli-1.0.0/CONTRIBUTING.md +37 -0
- lab_kit_cli-1.0.0/PKG-INFO +77 -0
- lab_kit_cli-1.0.0/README.md +59 -0
- lab_kit_cli-1.0.0/agents/reviewer.md +11 -0
- lab_kit_cli-1.0.0/docs/model.md +512 -0
- lab_kit_cli-1.0.0/method/METHOD.md +40 -0
- {lab_kit_cli-0.6.2 → lab_kit_cli-1.0.0}/pyproject.toml +8 -4
- lab_kit_cli-1.0.0/skills/lab-review/SKILL.md +57 -0
- lab_kit_cli-1.0.0/skills/lab-set-up/SKILL.md +47 -0
- lab_kit_cli-1.0.0/skills/lab-talk/SKILL.md +70 -0
- lab_kit_cli-1.0.0/skills/lab-work/SKILL.md +215 -0
- lab_kit_cli-1.0.0/src/lab_kit/__init__.py +2 -0
- lab_kit_cli-1.0.0/src/lab_kit/checks.py +321 -0
- lab_kit_cli-1.0.0/src/lab_kit/cli.py +191 -0
- lab_kit_cli-1.0.0/src/lab_kit/history.py +155 -0
- lab_kit_cli-1.0.0/src/lab_kit/record.py +945 -0
- lab_kit_cli-1.0.0/src/lab_kit/runs.py +412 -0
- lab_kit_cli-1.0.0/src/lab_kit/site.py +1156 -0
- lab_kit_cli-1.0.0/src/lab_kit/supervise.py +7 -0
- lab_kit_cli-1.0.0/src/lab_kit/views.py +236 -0
- lab_kit_cli-1.0.0/src/lab_kit/write.py +357 -0
- lab_kit_cli-1.0.0/tests/test_flow.py +395 -0
- lab_kit_cli-1.0.0/tests/test_integrity.py +864 -0
- lab_kit_cli-1.0.0/tests/test_refusals.py +146 -0
- lab_kit_cli-1.0.0/tests/test_site.py +308 -0
- lab_kit_cli-1.0.0/tools/import_run.py +61 -0
- lab_kit_cli-0.6.2/.github/workflows/ci.yml +0 -26
- lab_kit_cli-0.6.2/CHANGELOG.md +0 -58
- lab_kit_cli-0.6.2/CLAUDE.md +0 -1
- lab_kit_cli-0.6.2/CONTRIBUTING.md +0 -28
- lab_kit_cli-0.6.2/PKG-INFO +0 -114
- lab_kit_cli-0.6.2/README.md +0 -98
- lab_kit_cli-0.6.2/SETUP.md +0 -17
- lab_kit_cli-0.6.2/agents/reporter.md +0 -30
- lab_kit_cli-0.6.2/agents/reviewer.md +0 -36
- lab_kit_cli-0.6.2/agents/runner.md +0 -41
- lab_kit_cli-0.6.2/agents/scout.md +0 -28
- lab_kit_cli-0.6.2/docs/figures/how-lab-kit-works-dark.svg +0 -168
- lab_kit_cli-0.6.2/docs/figures/how-lab-kit-works-light.svg +0 -168
- lab_kit_cli-0.6.2/docs/figures/lab-loop-dark.svg +0 -159
- lab_kit_cli-0.6.2/docs/figures/lab-loop-light.svg +0 -159
- lab_kit_cli-0.6.2/docs/figures/number-chain-dark.svg +0 -150
- lab_kit_cli-0.6.2/docs/figures/number-chain-light.svg +0 -150
- lab_kit_cli-0.6.2/docs/spec/lab-model.md +0 -181
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/agents/reporter.md +0 -30
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/agents/reviewer.md +0 -36
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/agents/runner.md +0 -41
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/agents/scout.md +0 -28
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/address/SKILL.md +0 -75
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/configure/SKILL.md +0 -88
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/experiment/SKILL.md +0 -80
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/organise/SKILL.md +0 -92
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/plan-mission/SKILL.md +0 -80
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/publish/SKILL.md +0 -70
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/review/SKILL.md +0 -75
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/run/SKILL.md +0 -69
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/run-mission/SKILL.md +0 -84
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/set-up/SKILL.md +0 -72
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/set-up-lab/SKILL.md +0 -81
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.agents/skills/write/SKILL.md +0 -111
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.gitignore +0 -5
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.lab/frozen.sha256 +0 -2
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.lab/method/DISCIPLINE.md +0 -114
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.lab/method/LADDER.md +0 -73
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/.lab/migration.tsv +0 -4
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/AGENTS.md +0 -22
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/CLAUDE.md +0 -1
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/.folio/backlinks.json +0 -425
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/.folio/cards.json +0 -1
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/.folio/catalog.json +0 -491
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/.folio/journal.json +0 -143
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/.folio/nav.json +0 -1
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/.folio/search.json +0 -248
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/.gitignore +0 -4
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/assets/data/.gitkeep +0 -0
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/assets/figures/.gitkeep +0 -0
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/assets/refs.bib +0 -0
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/baseline.sha256 +0 -9
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/claims/C-1.md +0 -26
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/index.html +0 -19
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-concluded-the-pi-error-scaling-mission.md +0 -10
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-lab-set-up.md +0 -10
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-locked-the-pi-error-scaling-protocol.md +0 -10
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-opened-the-pi-error-scaling-mission.md +0 -10
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-r-1-fitted-the-mean-absolute-error.md +0 -10
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-ran-the-pi-error-scaling-protocol.md +0 -10
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-recorded-r-1.md +0 -9
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-recorded-r-2.md +0 -9
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-reported-the-pi-error-scaling-experiment.md +0 -9
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/journal/2026/2026-10-04-the-square-root-law-holds-per-size-predictions.md +0 -10
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/maps/research.html +0 -29
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/projects/lab/index.html +0 -31
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/protocols/pi-error-scaling.md +0 -66
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/questions/Q-1.md +0 -25
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/reports/pi-error-scaling-report/index.html +0 -47
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/results/R-1.md +0 -18
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/content/results/R-2.md +0 -19
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/docs/folio.yaml +0 -9
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/experiments/.gitkeep +0 -0
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/experiments/26-10-04-e_pi-error-scaling/bin/.gitkeep +0 -0
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/experiments/26-10-04-e_pi-error-scaling/bin/fold.py +0 -101
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/experiments/26-10-04-e_pi-error-scaling/bin/sample.py +0 -65
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/experiments/26-10-04-e_pi-error-scaling/lock.json +0 -11
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/experiments/26-10-04-e_pi-error-scaling/score.yaml +0 -50
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/lab.yaml +0 -5
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/missions/.gitkeep +0 -0
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/missions/26-10-04-m_pi-error-scaling.md +0 -42
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/ops/STATE.md +0 -7
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/ops/missions/.gitkeep +0 -0
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/runs/26-10-04-e_pi-error-scaling/26-10-04-2316-r_zlk7/out/estimates.tsv +0 -501
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/runs/26-10-04-e_pi-error-scaling/26-10-04-2316-r_zlk7/run.json +0 -43
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/substrate/README.md +0 -3
- lab_kit_cli-0.6.2/examples/monte-carlo-lab/substrate/estimator.py +0 -18
- lab_kit_cli-0.6.2/method/DISCIPLINE.md +0 -114
- lab_kit_cli-0.6.2/method/LADDER.md +0 -73
- lab_kit_cli-0.6.2/skills/experiment/SKILL.md +0 -80
- lab_kit_cli-0.6.2/skills/plan-mission/SKILL.md +0 -80
- lab_kit_cli-0.6.2/skills/review/SKILL.md +0 -75
- lab_kit_cli-0.6.2/skills/run-mission/SKILL.md +0 -84
- lab_kit_cli-0.6.2/skills/set-up-lab/SKILL.md +0 -81
- lab_kit_cli-0.6.2/src/lab_kit/__init__.py +0 -3
- lab_kit_cli-0.6.2/src/lab_kit/__main__.py +0 -3
- lab_kit_cli-0.6.2/src/lab_kit/budgets.py +0 -94
- lab_kit_cli-0.6.2/src/lab_kit/checks/__init__.py +0 -1
- lab_kit_cli-0.6.2/src/lab_kit/checks/base.py +0 -41
- lab_kit_cli-0.6.2/src/lab_kit/checks/files.py +0 -312
- lab_kit_cli-0.6.2/src/lab_kit/checks/locks.py +0 -156
- lab_kit_cli-0.6.2/src/lab_kit/checks/results.py +0 -188
- lab_kit_cli-0.6.2/src/lab_kit/checks/run.py +0 -63
- lab_kit_cli-0.6.2/src/lab_kit/cli.py +0 -233
- lab_kit_cli-0.6.2/src/lab_kit/commands/__init__.py +0 -1
- lab_kit_cli-0.6.2/src/lab_kit/commands/experiments.py +0 -97
- lab_kit_cli-0.6.2/src/lab_kit/commands/migrate.py +0 -682
- lab_kit_cli-0.6.2/src/lab_kit/commands/runs.py +0 -116
- lab_kit_cli-0.6.2/src/lab_kit/commands/setup.py +0 -150
- lab_kit_cli-0.6.2/src/lab_kit/commands/status.py +0 -41
- lab_kit_cli-0.6.2/src/lab_kit/data.py +0 -49
- lab_kit_cli-0.6.2/src/lab_kit/errors.py +0 -2
- lab_kit_cli-0.6.2/src/lab_kit/frozen.py +0 -48
- lab_kit_cli-0.6.2/src/lab_kit/gitlog.py +0 -55
- lab_kit_cli-0.6.2/src/lab_kit/lab.py +0 -173
- lab_kit_cli-0.6.2/src/lab_kit/ops.py +0 -193
- lab_kit_cli-0.6.2/src/lab_kit/protocol.py +0 -100
- lab_kit_cli-0.6.2/src/lab_kit/records.py +0 -357
- lab_kit_cli-0.6.2/src/lab_kit/rederive.py +0 -54
- lab_kit_cli-0.6.2/src/lab_kit/store.py +0 -273
- lab_kit_cli-0.6.2/src/lab_kit/supervise.py +0 -132
- lab_kit_cli-0.6.2/tests/conftest.py +0 -79
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/agents/reporter.md +0 -30
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/agents/reviewer.md +0 -36
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/agents/runner.md +0 -41
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/agents/scout.md +0 -28
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/address/SKILL.md +0 -75
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/configure/SKILL.md +0 -88
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/experiment/SKILL.md +0 -80
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/organise/SKILL.md +0 -92
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/plan-mission/SKILL.md +0 -79
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/publish/SKILL.md +0 -70
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/review/SKILL.md +0 -75
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/run/SKILL.md +0 -69
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/run-mission/SKILL.md +0 -83
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/set-up/SKILL.md +0 -72
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/set-up-lab/SKILL.md +0 -81
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.agents/skills/write/SKILL.md +0 -111
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.gitignore +0 -1
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.lab/frozen.sha256 +0 -2
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.lab/method/DISCIPLINE.md +0 -114
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/.lab/method/LADDER.md +0 -73
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/AGENTS.md +0 -22
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/.folio/backlinks.json +0 -425
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/.folio/cards.json +0 -1
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/.folio/catalog.json +0 -491
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/.folio/journal.json +0 -143
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/.folio/nav.json +0 -1
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/.folio/search.json +0 -248
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/.gitignore +0 -4
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/assets/data/.gitkeep +0 -0
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/assets/figures/.gitkeep +0 -0
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/assets/refs.bib +0 -0
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/claims/C-1.md +0 -26
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/index.html +0 -19
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-concluded-the-pi-error-scaling-mission.md +0 -10
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-lab-set-up.md +0 -10
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-locked-the-pi-error-scaling-protocol.md +0 -10
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-opened-the-pi-error-scaling-mission.md +0 -10
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-r-1-fitted-the-mean-absolute-error.md +0 -10
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-ran-the-pi-error-scaling-protocol.md +0 -10
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-recorded-r-1.md +0 -9
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-recorded-r-2.md +0 -9
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-reported-the-pi-error-scaling-experiment.md +0 -9
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/journal/2026/2026-10-04-the-square-root-law-holds-per-size-predictions.md +0 -10
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/maps/research.html +0 -29
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/projects/lab/index.html +0 -31
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/protocols/pi-error-scaling.md +0 -66
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/questions/Q-1.md +0 -25
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/reports/pi-error-scaling-report/index.html +0 -47
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/results/R-1.md +0 -18
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/content/results/R-2.md +0 -19
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/docs/folio.yaml +0 -8
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/.gitkeep +0 -0
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/.gitignore +0 -3
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/bin/.gitkeep +0 -0
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/bin/fold.py +0 -101
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/bin/sample.py +0 -65
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/lock.json +0 -6
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/runs/.gitkeep +0 -0
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/runs/20261004-231629/MANIFEST.sha256 +0 -1
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/runs/20261004-231629/out/estimates.tsv +0 -501
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/runs/20261004-231629/run.json +0 -31
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/experiments/pi-error-scaling/score.yaml +0 -50
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/lab.yaml +0 -5
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/ops/STATE.md +0 -7
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/ops/missions/.gitkeep +0 -0
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/ops/missions/2026-10-04-pi-error-scaling.md +0 -42
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/substrate/README.md +0 -3
- lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5/substrate/estimator.py +0 -18
- lab_kit_cli-0.6.2/tests/test_budgets.py +0 -213
- lab_kit_cli-0.6.2/tests/test_checks.py +0 -269
- lab_kit_cli-0.6.2/tests/test_end_to_end.py +0 -264
- lab_kit_cli-0.6.2/tests/test_example.py +0 -40
- lab_kit_cli-0.6.2/tests/test_instrument.py +0 -92
- lab_kit_cli-0.6.2/tests/test_migrate.py +0 -271
- lab_kit_cli-0.6.2/tests/test_runs.py +0 -78
- lab_kit_cli-0.6.2/tests/test_store.py +0 -115
- {lab_kit_cli-0.6.2 → lab_kit_cli-1.0.0}/.github/workflows/policy.yml +0 -0
- {lab_kit_cli-0.6.2 → lab_kit_cli-1.0.0}/.gitignore +0 -0
- {lab_kit_cli-0.6.2/tests/fixtures/monte-carlo-lab-0.5 → lab_kit_cli-1.0.0}/CLAUDE.md +0 -0
- {lab_kit_cli-0.6.2 → lab_kit_cli-1.0.0}/LICENSE +0 -0
|
@@ -0,0 +1,38 @@
|
|
|
1
|
+
name: gate
|
|
2
|
+
|
|
3
|
+
on:
|
|
4
|
+
push:
|
|
5
|
+
pull_request:
|
|
6
|
+
|
|
7
|
+
jobs:
|
|
8
|
+
check:
|
|
9
|
+
runs-on: ubuntu-latest
|
|
10
|
+
strategy:
|
|
11
|
+
matrix:
|
|
12
|
+
python: ["3.10", "3.12"]
|
|
13
|
+
steps:
|
|
14
|
+
- uses: actions/checkout@v4
|
|
15
|
+
- uses: actions/setup-python@v5
|
|
16
|
+
with:
|
|
17
|
+
python-version: ${{ matrix.python }}
|
|
18
|
+
# The site's tests need folio's views hook, in folio-kb 0.5.0 and later: the `site` extra brings it.
|
|
19
|
+
- name: Install lab-kit, and folio for the site's tests
|
|
20
|
+
run: |
|
|
21
|
+
python -m pip install --upgrade pip
|
|
22
|
+
python -m pip install -e ".[dev,site]"
|
|
23
|
+
- name: Tests
|
|
24
|
+
run: python -m pytest -q -rs
|
|
25
|
+
env:
|
|
26
|
+
LAB_SITE_TESTS: required # the site's tests run and fail like any other; they never skip here
|
|
27
|
+
# What a lab gets is the wheel: build it, install it alone, and set up a lab with it.
|
|
28
|
+
- name: The built wheel sets up a lab that passes its own check
|
|
29
|
+
if: matrix.python == '3.12'
|
|
30
|
+
run: |
|
|
31
|
+
python -m pip install build
|
|
32
|
+
python -m build --wheel
|
|
33
|
+
python -m venv /tmp/wheel
|
|
34
|
+
/tmp/wheel/bin/pip install dist/*.whl
|
|
35
|
+
mkdir /tmp/new-lab
|
|
36
|
+
cd /tmp/new-lab
|
|
37
|
+
/tmp/wheel/bin/lab init
|
|
38
|
+
/tmp/wheel/bin/lab check
|
|
@@ -21,16 +21,12 @@ jobs:
|
|
|
21
21
|
- uses: actions/setup-python@v5
|
|
22
22
|
with:
|
|
23
23
|
python-version: "3.12"
|
|
24
|
-
|
|
25
|
-
|
|
26
|
-
|
|
27
|
-
|
|
28
|
-
|
|
29
|
-
|
|
30
|
-
- run: python -m pytest -q
|
|
31
|
-
- name: Example lab
|
|
32
|
-
working-directory: examples/monte-carlo-lab
|
|
33
|
-
run: lab-kit check
|
|
24
|
+
# Never publish untested: the same gate CI runs, and here the site's tests must run too, so
|
|
25
|
+
# folio-kb 0.5.0 or later has to be on PyPI before lab-kit is released.
|
|
26
|
+
- run: python -m pip install -e ".[dev,site]"
|
|
27
|
+
- run: python -m pytest -q -rs
|
|
28
|
+
env:
|
|
29
|
+
LAB_SITE_TESTS: required
|
|
34
30
|
- name: Build
|
|
35
31
|
run: |
|
|
36
32
|
python -m pip install build
|
|
@@ -4,27 +4,25 @@ This file is the only copy of the operating rules for this checkout. `CLAUDE.md`
|
|
|
4
4
|
|
|
5
5
|
## Read
|
|
6
6
|
|
|
7
|
-
- `README.md`: what lab-kit is
|
|
8
|
-
- `docs/
|
|
9
|
-
- `method
|
|
10
|
-
- `skills/`: set-up
|
|
11
|
-
- `src/lab_kit/`: the `lab
|
|
12
|
-
- `examples/monte-carlo-lab/`: a small lab that passes the gate.
|
|
7
|
+
- `README.md`: what lab-kit is.
|
|
8
|
+
- `docs/model.md`: the contract. When code, a skill or the method disagrees with it, it wins.
|
|
9
|
+
- `method/METHOD.md`: the rules every lab keeps.
|
|
10
|
+
- `skills/`: lab-set-up, lab-work, lab-review, lab-talk. `agents/`: the reviewer.
|
|
11
|
+
- `src/lab_kit/`: the `lab` command. `record.py` reads a lab and works out every state, `views.py` shows it to the agent, `site.py` draws it for the operator inside folio, `write.py` and `runs.py` change it, `checks.py` checks it.
|
|
13
12
|
|
|
14
13
|
## The gate
|
|
15
14
|
|
|
16
15
|
```bash
|
|
17
16
|
python -m pytest -q
|
|
18
|
-
(cd examples/monte-carlo-lab && lab-kit check)
|
|
19
17
|
```
|
|
20
18
|
|
|
21
|
-
The tests
|
|
19
|
+
The tests build a small lab with the command itself and take it from a question to a reviewed result.
|
|
22
20
|
|
|
23
21
|
## Rules
|
|
24
22
|
|
|
25
|
-
-
|
|
26
|
-
- A
|
|
27
|
-
-
|
|
23
|
+
- The model wins. Change `docs/model.md` first, then the code, the method and the skills, in the same change.
|
|
24
|
+
- A rule a command holds is in `docs/model.md`, in the code and in a test together, with a case that it refuses.
|
|
25
|
+
- A command validates before it writes, and says what it wrote. No command writes a computed view into a lab.
|
|
28
26
|
- Never `git add -A`. Add the files you touched.
|
|
29
27
|
|
|
30
28
|
<!-- ops:lanes:start (generated from the owner's central rules; change them there, not here) -->
|
|
@@ -0,0 +1,159 @@
|
|
|
1
|
+
# Changelog
|
|
2
|
+
|
|
3
|
+
## 1.0.0
|
|
4
|
+
|
|
5
|
+
A rewrite from first principles. Nothing of 0.6 carries over: a lab kept with 0.6 is not read by 1.0,
|
|
6
|
+
and no command converts one. `docs/model.md` is the contract, and its section 8 sets the old model
|
|
7
|
+
beside the new.
|
|
8
|
+
|
|
9
|
+
What a lab is
|
|
10
|
+
|
|
11
|
+
- One tree. A goal holds questions. A question holds its steps, of three kinds: an experiment (a
|
|
12
|
+
comparison with its bars set in advance), a build (something an experiment needs that has to be
|
|
13
|
+
made first) and a reading (a look at what already exists, which suggests and never confirms). An
|
|
14
|
+
experiment holds its runs and its results. Every id spells the path: `G1.Q2.3`, and a result is
|
|
15
|
+
`G1.Q2.3:attack-success`.
|
|
16
|
+
- Decisions beside the tree. A decision is something discussed with the operator: what was seen, two
|
|
17
|
+
or three options, a recommendation, then what was decided and why, in the lab's words. Every goal
|
|
18
|
+
and question names the decision it came from.
|
|
19
|
+
- Cards. Each of these is one small Markdown file: front matter, a description, and named slots with
|
|
20
|
+
word budgets. A card never holds an id, a state or an outcome; all three are worked out from the
|
|
21
|
+
record when asked for. Nothing computed is kept in a lab.
|
|
22
|
+
- The files: `lab.toml`, `resources.toml`, `log.md`, `GLOSSARY.md`, `goals/`, `decisions/`, and
|
|
23
|
+
`runs/` outside git. Missions, protocols, attempts, budgets, ledgers and the folio documents that
|
|
24
|
+
held the record are gone.
|
|
25
|
+
|
|
26
|
+
How a number counts
|
|
27
|
+
|
|
28
|
+
- write the card, lock, run, result, review. Each step is a command that refuses when the one before
|
|
29
|
+
it is missing.
|
|
30
|
+
- An experiment says what it compares, each arm, its cases, and its confirmed-if and refuted-if
|
|
31
|
+
lines, each a sentence and the one result it tests (`attack-success < 0.05`). The outcome
|
|
32
|
+
(confirmed, refuted, unclear) is worked out from those lines and the results, never written.
|
|
33
|
+
- The lock holds what the experiment demands: the comparison, the arms, the cases, the lines, what
|
|
34
|
+
each outcome would mean, and the limits. That never changes. A changed demand, a failed results
|
|
35
|
+
review or a partial score is a new experiment that replaces the old one and says why. The lock
|
|
36
|
+
also records `plan/` as it was locked.
|
|
37
|
+
- An experiment is run until one run finishes. Each run pins the `plan/` folder it ran (the
|
|
38
|
+
configuration and the code, as files and no links). When `plan/` is not what the run before ran,
|
|
39
|
+
or for the first run what was locked, the run says what changed and why: `lab run --changed`.
|
|
40
|
+
The first run that finishes is the only source of results, nothing more runs under that
|
|
41
|
+
experiment, and its `plan/` stays as it ran.
|
|
42
|
+
- A run is detached, sealed when its command exits (every file's hash, and one over the record) and
|
|
43
|
+
moved into the store. A trial run comes before the lock and never gives a result.
|
|
44
|
+
- Every run is listed in `runs.tsv`, in git beside the lock: a row when it starts, a row with its
|
|
45
|
+
seal when it ends. The store is outside git, so this list is what says a run existed: a run
|
|
46
|
+
removed from the store, or a record rewritten, fails `lab check`, and a clone without the store
|
|
47
|
+
still knows what ran. A run that started and whose folder is gone for good is declared lost,
|
|
48
|
+
with the reason: `lab runs lost`.
|
|
49
|
+
- What git holds is tamper-evident from the commit that first holds it, not before: the lock is
|
|
50
|
+
committed when it is made and a run's row when it starts. `lab now` and `lab check` say what is
|
|
51
|
+
not committed yet.
|
|
52
|
+
- A result is a row in `results.tsv`: a name, a plain label, a value, how it reads, what it is
|
|
53
|
+
compared with, and the command that prints it again from the sealed run. `lab check` runs every
|
|
54
|
+
command again. A card names a result (`{attack-success}`) and never types its number.
|
|
55
|
+
- One review, after the results, by a session that did not do the work. A second, of the plan, only
|
|
56
|
+
before a run that uses a metered resource. A failed results review is final for that experiment.
|
|
57
|
+
|
|
58
|
+
Who decides
|
|
59
|
+
|
|
60
|
+
- The operator decides direction (the goals and the questions) and resources, in a discussion
|
|
61
|
+
recorded as a decision, not by approving requests. What a decision settles needs no further yes.
|
|
62
|
+
The agent decides everything else and records each choice in the log. A settled decision is
|
|
63
|
+
never reworded. A question is answered with `lab answer`, on a finished experiment.
|
|
64
|
+
- Resources are set once for the lab in `resources.toml`: what each is, where its key is kept, for a
|
|
65
|
+
metered one its unit, grant and amount for one run, and its rules as plain sentences. `lab run`
|
|
66
|
+
refuses a run that would pass the amounts; an experiment may carry an override, which is a
|
|
67
|
+
resource and a cap. A metered resource names the settled decision that set its amounts (`from`),
|
|
68
|
+
which lists it under `about:`, and an amount changes only together with it. A resource is never
|
|
69
|
+
removed, only retired, and use outside a run is a list that only grows.
|
|
70
|
+
|
|
71
|
+
The command, the method and the site
|
|
72
|
+
|
|
73
|
+
- One command, `lab` (`lab-kit` still starts it), with fifteen commands: `now`, `show`, `brief`,
|
|
74
|
+
`since` and `define` to look; `new`, `log`, `decide` and `answer` to write; `lock`, `run`, `runs`,
|
|
75
|
+
`review` and `check` for evidence; `init` once.
|
|
76
|
+
- One log, `log.md`, with five kinds of line: a decision recorded, a choice of the agent's, a
|
|
77
|
+
result, something stuck, and what the agent is starting on now (`doing`). The `doing` line is how
|
|
78
|
+
the operator sees that the lab is alive without asking.
|
|
79
|
+
- `lab check` holds what can be held mechanically: ids and links, card budgets and front matter,
|
|
80
|
+
a test on every line, unchanged locks, listed and sealed runs, results that re-derive and were
|
|
81
|
+
reviewed, use inside the resources' amounts, a decision behind every goal, question and amount,
|
|
82
|
+
the installed method and skills, and a history that only grows. It runs the commands in
|
|
83
|
+
`results.tsv`, so it is for a lab you trust. `--strict` fails when history or the store could not
|
|
84
|
+
be checked, which is what CI wants.
|
|
85
|
+
- A lab that is not written as the model says (a mistyped folder, a list written as one value, a
|
|
86
|
+
file that is not text) is reported by `lab check`, and stops no command and no page. Commands
|
|
87
|
+
validate everything before they write, and one writes at a time.
|
|
88
|
+
- The method is seven rules in one file, `method/METHOD.md`. Four skills (lab-set-up, lab-work,
|
|
89
|
+
lab-review, lab-talk) and one role (the reviewer). `lab init` installs them into a lab.
|
|
90
|
+
- The site is drawn by `lab_kit.site` inside folio, from the record, at each request. It needs
|
|
91
|
+
`folio-kb` 0.5.0 or later, as the optional extra `lab-kit-cli[site]`. The command itself depends
|
|
92
|
+
only on `pyyaml` (and `tomli` on Python 3.10), and needs Python 3.10 or later.
|
|
93
|
+
|
|
94
|
+
## 0.6.4
|
|
95
|
+
|
|
96
|
+
- `lab-append-only` accepts a recorded correction. `.lab/corrections.tsv` lists `path<TAB>commit<TAB>reason` for a commit that edited a mission log, a budget or a lock record in place and was later put right. That commit's version is left out of the file's history, so the versions before and after it must still show the file only growing, and the file's current version is never a listed one. The list is committed, so every waiver keeps its reason on the record.
|
|
97
|
+
- `lab-roster-frozen` compares a run's `config` with its protocol's pinned configuration as the run records it, through JSON. YAML reads a key such as `1:` as an integer and `run.json` holds it as the string `"1"`, so a run that used exactly its pinned configuration was reported as differing.
|
|
98
|
+
|
|
99
|
+
## 0.6.3
|
|
100
|
+
|
|
101
|
+
- A mission runs to its conclusion with no further input from the operator. The approval launches it: plan-mission hands it to run-mission at once, with no separate "go". A running mission never waits for an answer. A few lines it never crosses (new money, a credential, publishing outside the lab, a change to a frozen surface, anything irreversible outside git) unless its approved plan names the exact step; work that reaches one is recorded as blocked, and the mission carries on and concludes with it open for the next mission. plan-mission settles those lines in the plan instead of naming points where the run comes back to the operator.
|
|
102
|
+
- The commander settles what used to go back to the operator mid-mission: a reviewer disagreement (on the evidence, with a second reviewer), a protocol that must change after its lock (a new protocol), and a correction or retraction of a recorded result (a new record through the review skill, reported afterwards). Reports say what was decided, never ask. The reporter leaves pushing to the commander, and the reviewer's verdict no longer escalates past the commander. DISCIPLINE §8 and §14 and the spec say the same.
|
|
103
|
+
|
|
104
|
+
## 0.6.2
|
|
105
|
+
|
|
106
|
+
- The operator's words authorize; they are not the text of the record. The plan-mission skill writes a mission's objective in the lab's own words, as what it will establish inside a clear bound, and the operator approves that text; a vague prompt is interpreted, never pasted. The operator's sentence of approval is kept once, in one log line with its time. The run-mission skill keeps decisions in the log in the lab's words, never quoting the message that prompted them. DISCIPLINE §14 and the spec say the same. Missions already written are left as they are.
|
|
107
|
+
|
|
108
|
+
## 0.6.1
|
|
109
|
+
|
|
110
|
+
- `lab-kit migrate` rewrites a run's `config` with the same names it rewrites in its protocol's pinned configuration, so `lab-roster-frozen` holds after the move.
|
|
111
|
+
- The plan-mission skill points `ops/STATE.md` at `missions/`.
|
|
112
|
+
|
|
113
|
+
## 0.6.0
|
|
114
|
+
|
|
115
|
+
- One folder per run, outside git. A run is `runs/<experiment>/<YY-MM-DD-HHMM>-r_<4 characters>/` and holds everything it produced. lab-kit writes only `run.json` and `run.log` there; the rest is the experiment's own. `LAB_RUN` names the folder. `LAB_OUT`, `LAB_WORK` and `LAB_RUN_DIR` still work for one release.
|
|
116
|
+
- The seal. When the command exits, `run.json` records every file's sha256 in `files`, and one hash over the whole record in `seal`. `MANIFEST.sha256`, `work.sha256` and `store.json` are gone. `lab-evidence-sealed` checks a run's folder and its record against the seal.
|
|
117
|
+
- The store holds the runs. With `store:`, `runs/` is a committed link to `<store>/runs`. A run is written on local disk in `.lab/running/`, then moves into the store when it ends: copied, verified against its seal, renamed into place, made read-only, and only then deleted locally. `lab-kit publish --pending` retries a failed move, `lab-kit fetch` copies a run into `.lab/fetched/` on a machine without the store, and `lab-kit verify` checks a run against its seal. Where `runs/` does not resolve, as in CI, the gate skips the checks that read runs and says so once. `lab-work-stored` becomes `lab-run-stored`.
|
|
118
|
+
- Dated ids. Missions are `missions/<YY-MM-DD>-m_<slug>.md` (they were in `ops/missions/`), and experiments are `experiments/<YY-MM-DD>-e_<slug>/`. A protocol keeps its slug.
|
|
119
|
+
- `lab-kit migrate [--apply]` moves a lab from the 0.5.0 layout. It takes each experiment's date from its lock and each run's from its start. Each run's `out/`, `work/` and `run.log` go into one sealed folder in `runs/`. Every old path and id in git-tracked text is rewritten, and the lock hashes are taken again where a protocol's text changed with it. The rewritten documents are listed in the library's `baseline.sha256`, folio 0.3.6's `baseline`, so their permanence holds from the rewrite on, and every old name is kept in `.lab/migration.tsv`. Run it again after old runs land from machines on an older lab-kit.
|
|
120
|
+
- Run records are no longer append-only lab files in git: the seal holds them.
|
|
121
|
+
|
|
122
|
+
## 0.5.0
|
|
123
|
+
|
|
124
|
+
- lab-kit no longer runs a lab's code to check it. The `lab-selftest` check is gone, and with it the `tools:` key in `lab.yaml`: a lab that still has the key gets an error saying to delete it. lab-kit checks the record; the lab tests its code in its own test suite, which its gate runs beside `lab-kit check`.
|
|
125
|
+
- The method (DISCIPLINE §7) and the experiment and review skills say so: an experiment's command calls the lab's own tested code, `bin/` holds only what is particular to one experiment and may be empty, and a fold is tested on a planted case in the lab's suite. The lock still hashes `bin/` and records the commit, which says which version of the lab's code a run used.
|
|
126
|
+
- Locks and run records are unchanged.
|
|
127
|
+
|
|
128
|
+
## 0.4.0
|
|
129
|
+
|
|
130
|
+
- The run store. `store:` in `lab.yaml` names one shared filesystem, such as `/labs/<lab>`, where every finished run's raw output goes. When a run ends, `lab-kit run` publishes `work/` and `run.log` to the store at the run folder's own path: a copy into `.incoming/`, checked against `work.sha256` (nothing missing, extra or changed), renamed into place, made read-only. It then writes `store.json` beside `run.json` and deletes the local copy. A run in the store is never overwritten. A failed publish is noted in `run.log`, keeps the local copy and never changes the run's exit code.
|
|
131
|
+
- `lab-kit publish <slug>/<run id>... | --pending [--keep]`, `lab-kit fetch <slug>/<run id>` and `lab-kit verify <slug>/<run id>`. A machine that does not mount the store sets `LAB_STORE_HOST` to one that does (rsync over ssh, `sha256sum` there).
|
|
132
|
+
- `lab-work-stored` (warning): with a store set, every ended run with a `work.sha256` has a `store.json`.
|
|
133
|
+
- The experiment skill says to commit `store.json` with the run, and never to mount the store where the system under test runs.
|
|
134
|
+
- A lab without `store:` works as before.
|
|
135
|
+
|
|
136
|
+
## 0.3.0
|
|
137
|
+
|
|
138
|
+
- The lock covers the instrument. `lab-kit lock` records, beside the protocol's hash, `instrument` (one sha256 over the files in the experiment's `bin/`, sorted by path, each fed as its path and its bytes, `__pycache__` left out), `commit` (the git head) and `lab-kit` (the version). `lab-lock-intact` fails when `bin/` no longer hashes to `instrument`, and `lab-kit run` refuses to start on such a lock, as it already did for a changed protocol.
|
|
139
|
+
- Records name their commit. `lab-kit run` writes `commit` and `lab-kit` into `run.json` when it starts.
|
|
140
|
+
- Raw output is sealed. When the command exits, `lab-kit run` writes `work.sha256` beside `MANIFEST.sha256`, one line per file under `work/`, when that folder exists. `work/` stays out of git; its hashes go in, so a copy kept elsewhere can be verified.
|
|
141
|
+
- A lock or a run record without the new fields predates them and passes: records are never edited. The example lab's records are unchanged and are that test.
|
|
142
|
+
- The experiment skill says to commit `bin/` before the lock and to commit `work.sha256` with the run.
|
|
143
|
+
|
|
144
|
+
## 0.2.0
|
|
145
|
+
|
|
146
|
+
- Standing budgets. The operator can grant money once for many missions: `ops/budgets/<name>.yaml` lists grants, each with an amount and who granted it and when. A mission names the budget it draws on in `budget:`. `lab-mission-approved` checks that what the missions under a budget commit (an active mission's cap, or its spend when that is more; a concluded mission's spend) fits inside its grants; `lab-append-only` keeps grants append-only and the unit fixed. `lab-kit status` shows each budget's granted, committed and left amounts.
|
|
147
|
+
- A run record committed while its run was still going may be committed again once the run ends. `lab-append-only` accepts a later version of a `run.json` that differs from its first committed one only by `ended`, `exit` and `spent` filled in, each once; any other change, a closing field changed after it was set, and a deleted run record still fail.
|
|
148
|
+
- The method (DISCIPLINE §13, §14) and the plan-mission and run-mission skills: under a standing budget the commander may approve a mission that fits inside it and needs nothing that stays the operator's, decide the stops its own plan names and set mission caps inside the budget, recording each call in the mission log and reporting it afterwards. New money, credentials, publishing outside the lab and frozen surfaces stay the operator's.
|
|
149
|
+
|
|
150
|
+
## 0.1.0
|
|
151
|
+
|
|
152
|
+
The first release.
|
|
153
|
+
|
|
154
|
+
- The method: `DISCIPLINE.md` and `LADDER.md`, copied into each lab's `.lab/method/`.
|
|
155
|
+
- Five skills, set-up-lab, plan-mission, run-mission, experiment and review, and four agent roles, scout, runner, reviewer and reporter. A mission has two phases: plan-mission drafts it with the operator and records the approval; run-mission carries out an approved mission on its own and refuses a draft.
|
|
156
|
+
- The `lab-kit` command: `init`, `check`, `experiment`, `lock`, `run`, `runs`, `score`, `rederive`, `freeze`, `status` and `version`.
|
|
157
|
+
- The lab gate: `folio check`, then seventeen lab checks, in one report. `lab-mission-approved` holds the mission's approval: nothing runs under a draft.
|
|
158
|
+
- An example lab, `examples/monte-carlo-lab`, that passes the gate: does the error of a Monte Carlo estimate of pi shrink as 1/sqrt(n)?
|
|
159
|
+
- Built on folio 0.1.0 and its lab pack.
|
|
@@ -0,0 +1,37 @@
|
|
|
1
|
+
# Contributing
|
|
2
|
+
|
|
3
|
+
## Set up
|
|
4
|
+
|
|
5
|
+
```bash
|
|
6
|
+
python3 -m venv .venv
|
|
7
|
+
.venv/bin/pip install -e '.[dev,site]'
|
|
8
|
+
```
|
|
9
|
+
|
|
10
|
+
Python 3.10 or later. The `site` extra brings folio (`folio-kb` 0.5.0 or later), which only the
|
|
11
|
+
site's tests need; with a checkout of folio beside this one, `pip install -e ../folio` instead.
|
|
12
|
+
Without folio, `tests/test_site.py` skips itself and the rest runs; CI sets `LAB_SITE_TESTS=required`,
|
|
13
|
+
so there it never skips.
|
|
14
|
+
|
|
15
|
+
## The gate
|
|
16
|
+
|
|
17
|
+
```bash
|
|
18
|
+
python -m pytest -q
|
|
19
|
+
```
|
|
20
|
+
|
|
21
|
+
The tests build a small lab with the `lab` command itself and take it from a decision to a reviewed
|
|
22
|
+
result. CI runs them on Python 3.10 and 3.12, then builds the wheel, installs it alone and runs
|
|
23
|
+
`lab init` and `lab check` in an empty folder.
|
|
24
|
+
|
|
25
|
+
## How a change lands
|
|
26
|
+
|
|
27
|
+
- The contract is `docs/model.md`. Change it first, then the code, the method and the skills, in the
|
|
28
|
+
same change.
|
|
29
|
+
- A rule a command holds is in `docs/model.md`, in the code and in a test together, and the test has
|
|
30
|
+
a case that the command refuses.
|
|
31
|
+
- A command validates before it writes, and says what it wrote. No command writes a computed view
|
|
32
|
+
into a lab.
|
|
33
|
+
- The method, the skills and the reviewer role ship inside the package. A lab gets a new version of
|
|
34
|
+
them from `lab init`, never by hand, and `lab check` fails while they differ.
|
|
35
|
+
- Write plain words in short sentences. Name no particular project, person or organisation in the
|
|
36
|
+
docs or the code.
|
|
37
|
+
- Fail loud. A problem lab-kit cannot work around is an error with a message that says what to do.
|
|
@@ -0,0 +1,77 @@
|
|
|
1
|
+
Metadata-Version: 2.5
|
|
2
|
+
Name: lab-kit-cli
|
|
3
|
+
Version: 1.0.0
|
|
4
|
+
Summary: A research lab's record, its method, and the command that keeps both.
|
|
5
|
+
Project-URL: Homepage, https://github.com/pierg/lab-kit
|
|
6
|
+
Project-URL: Issues, https://github.com/pierg/lab-kit/issues
|
|
7
|
+
Author: Piergiuseppe Mallozzi
|
|
8
|
+
License-Expression: MIT
|
|
9
|
+
License-File: LICENSE
|
|
10
|
+
Requires-Python: >=3.10
|
|
11
|
+
Requires-Dist: pyyaml
|
|
12
|
+
Requires-Dist: tomli; python_version < '3.11'
|
|
13
|
+
Provides-Extra: dev
|
|
14
|
+
Requires-Dist: pytest; extra == 'dev'
|
|
15
|
+
Provides-Extra: site
|
|
16
|
+
Requires-Dist: folio-kb>=0.5.0; extra == 'site'
|
|
17
|
+
Description-Content-Type: text/markdown
|
|
18
|
+
|
|
19
|
+
# lab-kit
|
|
20
|
+
|
|
21
|
+
lab-kit keeps a research lab that coding agents work in. It fixes what the lab's record looks like,
|
|
22
|
+
the method that makes a number count, and one command, `lab`, that both keeps the rules and shows
|
|
23
|
+
the lab to an agent a level at a time. [folio](https://github.com/pierg/folio) serves the same
|
|
24
|
+
record to the operator as a live site: `lab_kit.site` draws the pages, and nothing is generated.
|
|
25
|
+
|
|
26
|
+
## Two readers
|
|
27
|
+
|
|
28
|
+
- **The operator** decides direction and resources, in conversation with the agent, and reads the
|
|
29
|
+
lab as a site. Nothing else reaches them.
|
|
30
|
+
- **The agent** plans, builds, runs and records, and decides everything else.
|
|
31
|
+
|
|
32
|
+
## One tree
|
|
33
|
+
|
|
34
|
+
```
|
|
35
|
+
Goal what the lab wants to be able to say
|
|
36
|
+
Question a yes-or-no question: what we expect, what would prove us wrong, what we know so far
|
|
37
|
+
Build something an experiment needs that has to be made first
|
|
38
|
+
Reading a look at what already exists: it suggests, and never confirms
|
|
39
|
+
Experiment what is compared, what would confirm, what would refute; then what was found
|
|
40
|
+
Run one try at producing what the experiment demands; the first that finishes counts
|
|
41
|
+
Result one number from that run, with the command that prints it again
|
|
42
|
+
Decision something discussed with the operator: what was seen, the options, what was decided and why
|
|
43
|
+
```
|
|
44
|
+
|
|
45
|
+
Every id spells the path: `G1.Q2.3`. Cards are small and written for the operator. Nothing computed is
|
|
46
|
+
kept: every state, and an experiment's outcome, is worked out from the record when it is asked for.
|
|
47
|
+
|
|
48
|
+
## How a number counts
|
|
49
|
+
|
|
50
|
+
```
|
|
51
|
+
write the card -> lock -> run -> result -> review -> the number may be used on a card
|
|
52
|
+
```
|
|
53
|
+
|
|
54
|
+
Each arrow is a command that refuses when the step before it is missing. An experiment's lines, each
|
|
55
|
+
with the result it tests, are locked before its run. The first run that finishes is the only source
|
|
56
|
+
of its results. The reviewer is a separate session that did not do the work. Every run is listed in a file
|
|
57
|
+
that git keeps, so one cannot be removed or rewritten quietly.
|
|
58
|
+
|
|
59
|
+
## Install
|
|
60
|
+
|
|
61
|
+
```text
|
|
62
|
+
uv tool install lab-kit-cli
|
|
63
|
+
lab init
|
|
64
|
+
```
|
|
65
|
+
|
|
66
|
+
Then ask your agent to follow `.agents/skills/lab-set-up/SKILL.md`. `lab --help` lists the fifteen
|
|
67
|
+
commands.
|
|
68
|
+
|
|
69
|
+
The command needs Python 3.10 or later and nothing of folio. The site does: install `lab-kit-cli[site]`
|
|
70
|
+
where folio serves the lab, which brings `folio-kb` 0.5.0 or later, and list `lab_kit.site` under
|
|
71
|
+
`views:` in the charter of the lab's `docs/` library.
|
|
72
|
+
|
|
73
|
+
## Read more
|
|
74
|
+
|
|
75
|
+
- [docs/model.md](https://github.com/pierg/lab-kit/blob/main/docs/model.md): the contract. What you read, how a number becomes a result, who decides, the commands, the files.
|
|
76
|
+
- [method/METHOD.md](https://github.com/pierg/lab-kit/blob/main/method/METHOD.md): the seven rules every lab keeps.
|
|
77
|
+
- [skills/](https://github.com/pierg/lab-kit/blob/main/skills/): lab-set-up, lab-work, lab-review and lab-talk. [agents/](https://github.com/pierg/lab-kit/blob/main/agents/): the reviewer.
|
|
@@ -0,0 +1,59 @@
|
|
|
1
|
+
# lab-kit
|
|
2
|
+
|
|
3
|
+
lab-kit keeps a research lab that coding agents work in. It fixes what the lab's record looks like,
|
|
4
|
+
the method that makes a number count, and one command, `lab`, that both keeps the rules and shows
|
|
5
|
+
the lab to an agent a level at a time. [folio](https://github.com/pierg/folio) serves the same
|
|
6
|
+
record to the operator as a live site: `lab_kit.site` draws the pages, and nothing is generated.
|
|
7
|
+
|
|
8
|
+
## Two readers
|
|
9
|
+
|
|
10
|
+
- **The operator** decides direction and resources, in conversation with the agent, and reads the
|
|
11
|
+
lab as a site. Nothing else reaches them.
|
|
12
|
+
- **The agent** plans, builds, runs and records, and decides everything else.
|
|
13
|
+
|
|
14
|
+
## One tree
|
|
15
|
+
|
|
16
|
+
```
|
|
17
|
+
Goal what the lab wants to be able to say
|
|
18
|
+
Question a yes-or-no question: what we expect, what would prove us wrong, what we know so far
|
|
19
|
+
Build something an experiment needs that has to be made first
|
|
20
|
+
Reading a look at what already exists: it suggests, and never confirms
|
|
21
|
+
Experiment what is compared, what would confirm, what would refute; then what was found
|
|
22
|
+
Run one try at producing what the experiment demands; the first that finishes counts
|
|
23
|
+
Result one number from that run, with the command that prints it again
|
|
24
|
+
Decision something discussed with the operator: what was seen, the options, what was decided and why
|
|
25
|
+
```
|
|
26
|
+
|
|
27
|
+
Every id spells the path: `G1.Q2.3`. Cards are small and written for the operator. Nothing computed is
|
|
28
|
+
kept: every state, and an experiment's outcome, is worked out from the record when it is asked for.
|
|
29
|
+
|
|
30
|
+
## How a number counts
|
|
31
|
+
|
|
32
|
+
```
|
|
33
|
+
write the card -> lock -> run -> result -> review -> the number may be used on a card
|
|
34
|
+
```
|
|
35
|
+
|
|
36
|
+
Each arrow is a command that refuses when the step before it is missing. An experiment's lines, each
|
|
37
|
+
with the result it tests, are locked before its run. The first run that finishes is the only source
|
|
38
|
+
of its results. The reviewer is a separate session that did not do the work. Every run is listed in a file
|
|
39
|
+
that git keeps, so one cannot be removed or rewritten quietly.
|
|
40
|
+
|
|
41
|
+
## Install
|
|
42
|
+
|
|
43
|
+
```text
|
|
44
|
+
uv tool install lab-kit-cli
|
|
45
|
+
lab init
|
|
46
|
+
```
|
|
47
|
+
|
|
48
|
+
Then ask your agent to follow `.agents/skills/lab-set-up/SKILL.md`. `lab --help` lists the fifteen
|
|
49
|
+
commands.
|
|
50
|
+
|
|
51
|
+
The command needs Python 3.10 or later and nothing of folio. The site does: install `lab-kit-cli[site]`
|
|
52
|
+
where folio serves the lab, which brings `folio-kb` 0.5.0 or later, and list `lab_kit.site` under
|
|
53
|
+
`views:` in the charter of the lab's `docs/` library.
|
|
54
|
+
|
|
55
|
+
## Read more
|
|
56
|
+
|
|
57
|
+
- [docs/model.md](docs/model.md): the contract. What you read, how a number becomes a result, who decides, the commands, the files.
|
|
58
|
+
- [method/METHOD.md](method/METHOD.md): the seven rules every lab keeps.
|
|
59
|
+
- [skills/](skills/): lab-set-up, lab-work, lab-review and lab-talk. [agents/](agents/): the reviewer.
|
|
@@ -0,0 +1,11 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: reviewer
|
|
3
|
+
description: >-
|
|
4
|
+
Reads one experiment's plan or results with fresh context and records a verdict. Use for every
|
|
5
|
+
results review, and for a plan review before a run that uses a metered resource. Give it only
|
|
6
|
+
the experiment's id and the stage. Never use it in the session that did the work.
|
|
7
|
+
---
|
|
8
|
+
|
|
9
|
+
You review one experiment. You did not plan it, build it or run it, and you were told nothing about
|
|
10
|
+
it except its id and the stage. Follow `.agents/skills/lab-review/SKILL.md`. Write nothing in the lab
|
|
11
|
+
except your verdict, through `lab review`.
|