Proof Matrix¶
Page Maps¶
graph LR
family["Python Programming"]
program["Python Meta-Programming"]
section["Guides"]
page["Proof Matrix"]
capstone["Capstone evidence"]
family --> program --> section --> page
page -.applies in.-> capstone
flowchart LR
orient["Orient on the page map"] --> read["Read the main claim and examples"]
read --> inspect["Inspect the related code, proof, or capstone surface"]
inspect --> verify["Run or review the verification path"]
verify --> apply["Apply the idea back to the module and capstone"]
Use this page when the claim is already clear and you want the fastest honest route to evidence.
This page is different from the proof ladder. The ladder helps you choose how much proof you need. The matrix helps you choose where to go once you know what claim you are trying to settle.
How to use this page¶
- State the claim in one sentence.
- Find the row that matches it most closely.
- Start with the first command listed there.
- Inspect the first surface before escalating.
If you still do not know what claim you are testing, go back to Proof Ladder or Pressure Routes.
Core runtime claims¶
| Claim | Best first command | Best first evidence surface |
|---|---|---|
| public runtime shape stays observable without executing plugin work | make PROGRAM=python-programming/python-meta-programming capstone-manifest |
manifest.json plus cli.py |
| registered plugins stay deterministic and inspectable | make PROGRAM=python-programming/python-meta-programming capstone-registry |
registry.json, framework.py, and tests/test_registry.py |
| one plugin contract remains legible from the public CLI | make PROGRAM=python-programming/python-meta-programming capstone-plugin |
plugin.json plus plugins.py |
| one field invariant stays attached to attribute ownership | make PROGRAM=python-programming/python-meta-programming capstone-field |
field.json, fields.py, and tests/test_fields.py |
| one field's descriptor role, storage owner, and collection path remain observable without construction | make PROGRAM=python-programming/python-meta-programming capstone-field-ownership |
field-ownership.json, framework.py, and field ownership tests |
| the complete field system names accepted and rejected powers without construction | make PROGRAM=python-programming/python-meta-programming capstone-field-system |
field-system.json, framework.py, and field-system tests |
| action decorators preserve public callable shape while adding behavior | make PROGRAM=python-programming/python-meta-programming capstone-action |
action.json, actions.py, and tests/test_runtime.py |
| action values match the supported runtime hint subset without hidden execution | make PROGRAM=python-programming/python-meta-programming capstone-check-action |
check-action.json, actions.py, and runtime tests |
| class-creation provenance separates declared, inherited, and generated surfaces | make PROGRAM=python-programming/python-meta-programming capstone-class-creation |
class-creation.json, framework.py, and runtime tests |
| the capstone publishes accepted, constrained, and rejected powers without construction or execution | make PROGRAM=python-programming/python-meta-programming capstone-governance |
governance.json, governance.py, and governance/CLI tests |
| generated constructor and action signatures still reflect declared fields honestly | make PROGRAM=python-programming/python-meta-programming capstone-signatures |
signatures.json, framework.py, and runtime tests |
| one concrete runtime action remains traceable after execution | make PROGRAM=python-programming/python-meta-programming capstone-trace |
trace.json plus runtime tests |
Bundle and review claims¶
| Claim | Best first command | Best first evidence surface |
|---|---|---|
| you can inspect the runtime without reading source first | make PROGRAM=python-programming/python-meta-programming inspect |
the saved inspection bundle under artifacts/inspect/... |
| you can follow a guided repository walkthrough honestly | make PROGRAM=python-programming/python-meta-programming capstone-walkthrough |
the walkthrough bundle under artifacts/tour/... |
| executable tests and public review surfaces still agree | make PROGRAM=python-programming/python-meta-programming capstone-verify-report |
pytest.txt plus matching public outputs |
| the raw executable suite still passes | make PROGRAM=python-programming/python-meta-programming test |
the core test files |
| the strongest local confirmation route still holds | make PROGRAM=python-programming/python-meta-programming capstone-confirm |
the confirmation route plus its public command surfaces |
| the full public proof route still builds | make PROGRAM=python-programming/python-meta-programming proof |
inspection, walkthrough, and verification bundles together |
Course outcome to evidence route¶
| If you are trying to prove... | Best first route | Best first surface |
|---|---|---|
| I can inspect runtime structure safely | capstone-manifest, capstone-registry, or inspect |
public CLI outputs plus framework.py |
| wrappers preserved metadata and public callable shape | capstone-action, capstone-signatures, or capstone-trace |
actions.py, runtime tests, and trace output |
| field or descriptor ownership is placed honestly | descriptor-lookup-lab, then capstone-field-ownership |
lab evidence, framework.py, fields.py, and field ownership tests |
| field-system power stops at an honest architecture boundary | descriptor-system-lab, then capstone-field-system |
system evidence, framework.py, and field-system tests |
| class-creation control is justified and deterministic | class-creation-lab, then capstone-class-creation |
hook and owner evidence, framework.py, and class-creation tests |
| runtime-governance claims stop at their executable boundaries | runtime-governance-lab-test, then runtime-governance-lab |
lab JSON, five owning modules, and focused governance tests |
| review routes remain observational rather than magical | capstone-governance, then inspect |
governance JSON, saved bundle manifest, CLI, and governance tests |
Module bridge¶
| Module range | Main question | Best first evidence route |
|---|---|---|
| Modules 01 to 03 | what can be observed safely without accidental execution | manifest, registry, signatures, or inspect |
| Modules 04 to 05 | did a wrapper preserve evidence and keep its policy claims bounded | action-wrapper, check-action, trace, and tests/test_runtime.py |
| Module 06 | what can post-construction tools own and what requires class-creation timing | class-customization-lab, class-creation, and runtime tests |
| Module 07 | can protocol status, lookup precedence, binding, and storage ownership be predicted and observed | descriptor-lookup-lab, field-ownership, and descriptor lab tests |
| Module 08 | do cache, backend, composition, hint, and record boundaries stay observable without becoming hidden architecture | descriptor-system-lab, field-system, and descriptor-system tests |
| Module 09 | do resolution, declaration, shaping, and registration have the smallest honest owners | class-creation-lab, class-creation, and focused creation tests |
| Module 10 | can trust, contract, rollback, tooling, and escalation decisions be proved before capstone transfer | runtime-governance-lab-test, runtime-governance-lab, then capstone-governance |
Quick review questions¶
| If you are asking... | Best first move |
|---|---|
| Where should I start if the public runtime shape already feels suspicious? | capstone-manifest or capstone-registry |
| Where should I start if the issue is descriptor ownership? | descriptor-lookup-lab, then capstone-field-ownership |
| Where should I start if several field powers are becoming architecture? | descriptor-system-lab, then capstone-field-system |
| Where should I start if registration is being used to justify a metaclass? | class-creation-lab, then compare __init_subclass__ with capstone-class-creation |
| Where should I start if a dynamic mechanism's safety or enforcement claim feels too large? | runtime-governance-lab-test, then inspect the matching evidence key and focused test |
| Where should I start if the capstone needs a power decision rather than another runtime trace? | capstone-governance, then the mechanism-specific observational command |
| Where should I start if the issue is one exported field or action contract? | capstone-field or capstone-action |
| Which saved bundle is strongest for a human review? | inspect or capstone-verify-report |
| Which route should I use before approving a larger change? | capstone-verify-report or proof |